Home

SPSD/M - Data Library

1. Min 5000 235 1 324 0 9777 7 1400 5001 10000 215 4 246 0 801 6 152 10001 15000 325 3 368 2 756 7 192 15001 20000 187 6 285 9 715 1 177 20001 25000 130 1 209 5 886 7 198 25001 30000 55 6 81 1 976 3 164 30001 35000 17 9 33 8 406 1 79 35001 40000 7 1 9 7 480 9 74 40001 45000 7 0 7 6 241 3 40 45001 Max 23 6 34 1 579 5 72 LE o oM All 1204 7 1600 0 14622 0 2547 LE o MM 4 oM Table 3U shows the distribution of EITC benefits over individuals by disposable income group Benefits are paid to persons with high incomes because sources of income other than from employment are not being taken into account in determining benefits Distribution report Earned income tax credit for Individuals Total observations 7648 Zero observations 6656 The following statistics are based on 992 non zero observations Descriptive Statistics Sum of weights 1600014 Weighted Sum 1204740412 Weighted Sum of Squares 1133173105452 Weighted Mean 753 Extreme Values with associated household numbers Minima hdseqhh Maxima hdseghh 1 2885 1200 56 1 2885 1200 57 3 352 1200 58 3 352 1200 59 6 2871 1200 66 Selected Quantiles Q1 445 P1 43 P90 1200 P20 437 P60 939 Introduction and Overview Guide Page 79 SPSD M Version 9 2 Med 735 P2 44 P95 1200 P30 497 P70 1062 Q3 1200 P5 87 P98 1200 P40
2. Min 5000 639 1 56 32 9 7 5001 10000 648 1 48 1573 1 10001 15000 771 1 65 272 6 53 15001 20000 616 1 38 173 3 34 20001 25000 640 1 61 2380 47 25001 30000 719 1 85 385 4 66 30001 35000 631 1 84 497 7 67 35001 40000 559 1 92 399 9 66 40001 45000 551 1 87 328 5 61 45001 Max 510 2 14 1321 1 206 oM Al 594 1 89 3806 8 624 MM Examining Table 2U we note that the distribution of the mean change in disposable income in the first column does not decrease monotonically with increasing income The fluctuations can be explained by the number of children for whom benefits were paid Note also that while there are only ten income groups the underlying sample size is 7 for the Min 5 000 income group In general attention to sample size is required as the number of records in each cell of a table can get small very quickly as the total number of cells increases Had this example been run with the full sample however the record counts would have been about twenty times those as shown in Table 2U SESSION 3 CHILD BENEFIT TRIAL SIMULATION This example which is included here for purely illustrative purposes shows the steps that a user would take to perform the following analysis 1 Eliminate Federal Family Allowances 2
3. Introduction and Overview Guide Page 45 SPSD M Version 9 2 14 Do you wish to modify any control parameters gt y 15 gt cprdesc Description of SPSM run string 16 CPRDESC gt Tax transfer 1992 Population 1992 Incomes 1992 1992 5 gt outmrsflag Variant results file creation flag 17 OUTMRSFLAG gt 1 gt outmrsvars Variant results file variables string 18 OUTMRSVARS gt 00 immicons immdisp imffa impfa 19 gt go No files will be overwritten with these control parameters 20 Any further control parameter changes gt n Writing out control parameters to test1 cpr 14 The parameters which control the reading of input files and production and contents of output results files are control parameters 15 The control parameter file description should be changed to add the fact that a results file is being produced 16 The new description has the same information as the old but it notes that a results file will be created 17 This parameter should be set to one to cause a results file to be produced 18 This parameter is set to a character string whose contents are the names of valid SPSM variant variables which will be written into the results file This setting specifies four variables to be written to the results file disposable income consumable income federal family allowances and provincial family allowances This setting will replace the default of immicons 1
4. SPSD M Introduction and Overview Guide This guide is designed to give an overview and introduction to the SPSD M It describes the construction of an SPSM simulation run and contains several examples of model use It should be read completely by anyone who will be using the SPSD M for research purposes Statistics Statistique l ivi Canada Canada Canada Table of Contents E CCU ENC HOTE o Sass et alae ao pe re cate ss cee ans ore ees Cases Soe I aes Soe I E 1 What You Should Already Know ccccccceeeeeeeeeeeeeeeeeeeeeeeeeeeenneeeeeeeeeeeeeees 1 Using This UIGG eis recien nk Sd CN etd eld eM ind ud depo ducimus e 2 Overview of the SP SIM Jia isso rotam ese ven cans nen deneniendvasemeudtecenies 2 Capabilities and Fealures 6 ire eoe ccs xe daas Ht are lat ua neat 3 SotbWare arnd Faellillas bodie bh e bt bo ebd 5 GUSTS AIT DEN RM DP ain DRM OUMMPE DM 6 Operating System Considerations ccccceccceceeeeeeeeeeeeneeeeeeeeeeeeeeeeeaaeees 7 lassi ce 9 Example Simulations nace ab To a nacta tanta ana 11 Examiblrig OUIDUE ches teshethe diese Ah ed eh Lt heen LEA CAL DA AAA EAR AL EAD AAA 15 Understanding Basic Components ccccceeeeeeeeeeceeeeeeeeeeeeeeeaeeaaeeeeeeeeeeeeeeees 16 DENION C Cetera tinesettosislenadernateataletiiataiainctatet T 17 Level of Analysis cien iaces tent te teh eode piae Made need e tanh ceed heated Ace 20 AAS WTS ceste etc al lh Ola dns 21 SIMULA TOT oem 21 Database A
5. TH Purpose This file contains the necessary control parameters to run HH tutorial session 2 The explanation of the parameters may HH be found in the Introduction and Overview Guide The User is still required to edit the tax transfer parameters in HH the user dialogue TH 2 THHHHHBHHHHHHHBHHHHHHHHHHHHHHHHHHHHHBHHHHHHHHHHHHBHHHHHBHHHHHEE IEHEHHHBHHNE 2 1 Model Control Parameters CPRDESC Base Variant elimination of Family Allowances X tab Table 2 1 7 Record selection facility SELFLAG 1 Selection facility activation flag 3 SELUNIT 2 Selection facility family level 4 SELSPEC nfnkids gt 0 Selection specification HH 2 1 5 Base Information BASDESC My 1992 Base Results File 1992 SPSD M Defaults in Place 5 BASMETH 1 Method of creating base variables 6 INPBASMRS test1 mrs Name of base results file in 1 Descriptive headers which document the file may be found on all include cpi mpi files supplied with the SPSD M 2 Correct file documentation should be maintained The SPSM will interpret any value supplied here as a character string up to a maximum of 60 characters The SPSM will truncate any specified strings which are longer than 60 characters 3 The selection unit specifies that selection will occur at the census family level of analysis 4 This expression will be evaluated for every individual on the database and then propagated to the nuclear family level of analysis 5 Base results wil
6. Name of FAMEX vector file in INPWGT SPSD v90y92t wgt Name of weight file in Growth INPAPR SPSD ba96 92 apr 3t Name of database adjustment parameter File Scenario Control no base variant computed using parms gt BASMETH 0 3t Method of creating base variables VARMETH 2 Method of creating variant variables INPVARMPR SPSD ba92 mpr 3t Name of variant tax transfer parameter File Output Files OUTMRSFLAG 0 Variant results file creation flag OUTCPR test1 cpr Name of control parameter file out Control parameters showing all input and output file specifications are displayed The description of the control file reflects the status quo 1992 tax transfer system with sales taxes The 596 database files are used as input which will provide a 5 sample Default Table O will be output The BASMETH parameter method of calculating base results is set to zero by default so that base results will not be calculated VARMETH method of calculating variant results is set to two by default so that the standard algorithm will be used to calculate variant results Note that the INPVARMPR parameter is set so that the variant tax transfer algorithm control parameters will be read from the default file SPSD BA92 MPR Note that the flag for controlling the writing of a results file is turned off We will have to edit the value of this parameter during the dialogue in order to activate the facility and cause a results file to be produced
7. 0 0 1593 2 6 8 1600 0 Unaffected by EITC Flag 000 8154 9 6439 6 27 5 14622 0 TooMMMMMMMMMMeeeey TM To Table 1U seems to indicate that the program has been correctly implemented as specified No benefits are paid to children aged 20 and under Total benefits of 1 2 billion accrue to 1 6 million individuals There are only 34 300 individuals over the age of 64 This is because the selection facility restricts the sample to census families having children under age 21 In total only an estimated 16 222 000 1600 0 14622 0 Canadians of any age live in such families in 1992 Table 2U Selected Quantities for Individuals by Wages amp salaries Group Wages amp Earned income Average Received EITC Unaffected by salaries tax credit Benefits Flag Gainer EITC Flag Group M 000 000 oM Min 0 297 3 832 0381 357 3 87434 1 8000 356 1 693 1384 513 7 1592 9 8001 12000 317 6 1041 1556 305 1 308 2 12001 24000 233 7 551 3439 4239 8259 24001 Max 0 0 0 0000 0 0 3151 6 oM o oM oM 4o Al 12047 752 9561 1600 0 14622 0 MM Table 2U shows the same results as Table 1U except they are now broken out by individual employment earnings level As we would expect
8. 22222222222 2 Database Adjustment Parameters 15 Loading database adjustment parameters from SPSD ba97_92 apr Current values of selected database adjustment parameters GFISENF 10 Growth Factor Self employed income non 0 74940 0 89140 0 85640 0 79350 0 87880 0 77390 0 79200 0 75870 0 79820 0 85950 GFIEMP 10 Growth Factor Employment income 0 96450 0 96450 0 99370 0 99370 0 99370 0 99370 0 96230 0 96230 0 95560 0 95560 0 94580 0 94580 0 97530 0 97530 0 88710 0 88710 0 90040 0 90040 0 95650 0 95650 Do you wish to modify any database adjustment parameters gt n 2l l2llll lllll l Tax Transfer Parameters Joading base tax transfer parameters from SPSD ba92 mpr Joading variant tax transfer parameters from SPSD ba92 mpr current values of selected variant tax transfer parameters MPRDESC Current values for 1992 CTFLAG 1 Commodity tax activation flag CTOPT 2 Commodity tax calculation method CTDFLAG 0 Commodity tax detailed calculation flag TARGETYEAR 1992 Year of analysis 16 Do you wish to modify any variant tax transfer parameters gt y 12 We will create a table which specifies the exact information we need To do this we activate the X tab facility by setting this parameter to 1 13 Itis good practice to indent each line of a multi line parameter using a tab or spaces This makes the parameter much easier to
9. 608 P80 1200 P10 236 P99 1200 Histogram Plot I hu el quee B MM sees P1 P5 P10 P20 P30 P40 P50 P60 P70 P99 43 87 236 437 497 608 735 939 1062 1200 The distribution report indicates that the program has been implemented correctly The maximum individual benefit is 1 200 as expected Guide to SPSD M Documentation SPSD M INTRODUCTORY MANUAL The SPSD M is a complex and sophisticated tool for the analysis of the intricacies of Canada s tax transfer system The Introductory Manual is designed to get the user up and running in a minimum amount of time while providing an overview of the elements of the SPSD M which must be understood to conduct valid research and analysis The SPSD M Introductory Manual consists of three documents Installation Guide This guide describes machine requirements for different configurations of SPSD M and describes the installation on the user s PC Lists of all files supplied with the SPSD M can be found in this guide Introduction and Overview This is the current guide which is designed to give an overview and introduction to the SPSD M It describes the construction of an SPSM simulation run and also contains several worked examples of model use Much of the material presented here is
10. 9 0 Database Adjustment Algorithm Standard adjustment Standard Tax Transfer Algorithm Version 9 0 88 05 Alternate Tax Transfer Algorithm none aucun Compiler Operating System MSC 12 0 WIN NT Copyright c Minister of Industry 1984 2001 Press F1 for context specific help Press Ctrl Break to terminate LLLLLlLlLlLllllll22lll 2 2 Control Parameters 2 Enter name of input control parameter file gt spsd ba92t cpr Loading control parameters from spsd ba92t cpr 3 Enter specification for generating output files gt factc1 Generating default output file names 1 The current directory should be SPSMTEST and the SPSM is invoked as usual 2 We shall start by reading the 1992 default parameter file 3 As discussed earlier the name of this first program run will be FACTC1 Current values of selected control parameters CPRDESC Tax ransfer 1992 Population 1992 Incomes 1992 1992 5 LICENSEE Statistics Canada SPSD M licensee Selection Sample SELFLAG 0 Selection facility activation flag SAMPLEREQ _ 1 000000000 Size of sample requested Input Databases INPSPD SPSD v90y97t spd Name of SPSD file in INPFXV SPSD v90y97t fxv Name of FAMEX vector file in INPWGT SPSD v90y92t wgt Name of weight file in Growth INPAPR SPSD ba97 92 apr Name of database adjustment parameter file Scenario Control no base variant computed using parms BASMETH 0 3t
11. 9 0 Database Adjustment Algorithm Standard adjustment Standard Tax Transfer Algorithm Version 9 0 88 05 Alternate Tax Transfer Algorithm none aucun Compiler Operating System MSC 12 0 WIN NT Copyright c Minister of Industry 1984 2001 Press F1 for context specific help Press Ctrl Break to terminate Introduction and Overview Guide Page 67 SPSD M Version 9 2 Enter name of input control parameter file gt spsd ba92t cpr Loading control parameters from spsd ba92t cpr Enter specification for generating output files gt factc3 Generating default output file names The FACTCS run will start with the default control parameter file for 1992 The filenames of the different types of output file will be FACTC3 37 37 38 39 40 41 42 38 39 40 Current values of selected control parameters CPRDESC Tax ransfer 1992 Population 1992 Incomes 1992 1992 5 LICENSEE Statistics Canada SPSD M licensee Selection Sample SELFLAG 0 Selection facility activation flag SAMPLEREQ _ 1 000000000 Size of sample requested Input Databases INPSPD SPSD v90y97t spd Name of SPSD file in INPFXV SPSD v90y97t fxv Name of FAMEX vector file in INPWGT SPSD v90y92t wgt Name of weight file in Growth INPAPR SPSD ba97 92 apr Name of database adjustment Parameter file Scenario Control no base variant computed using parms BASMETH 0 3t Method of creating base va
12. EXAMINING OUTPUT As mentioned earlier all output from the SPSM is written to files Some of these files are designed to be read by other computer programs Others are meant to be examined either on the screen or on paper There are various ways to do this The simplest way is to display the results on the screen using the MS DOS type command eg type firstime tbl Another way is to send the file to a printer using the MS DOS print or copy command The problem of course is that the text will scroll off the screen if it is more than 23 lines long and it will wrap around if it is more than 80 characters wide Since many default tables are 132 characters wide a printer must either have wide paper or use a condensed font A text editor provides a good means for viewing output because the user can then easily scroll back and forth between the desired portion of the output By importing a file into a spreadsheet the user can not only move around the output easily but can also perform subsequent calculations on the results This technique is used in the spreadsheet interface facilities described in the Tools User s Guide The following is a listing of the output produced in our sample run above The output is found in the file named FIRSTIME TBL This is only the first of ten possible default tables and is called Table 0 The format and contents of this table are fixed Users may not change either the rows or the columns of the table Some de
13. Increase the Child Tax Credit basic amount to leave total consumable income unchanged revenue neutrality 3 Assess the effects on the Federal and Provincial Treasuries 4 Assess the effects in terms of progressivity The first task in performing this analysis is to locate the tax transfer algorithm Introduction and Overview Guide Page 56 SPSD M Version 9 2 parameters relevant to family allowances and the child tax credit This is normally done by consulting the Parameter Guide However the parameters are also listed by category in Appendix A of the printed document The family allowance parameters are found in the government transfer category and look as follows Family Allowance FAFLAG Family allowance flag PFACFLAG Use provincial configuration flag All Provinces Except Alberta and Qu bec STDFA Standard federal family allowance per child Alberta AFAC1 Alberta FA benefit per child aged 0 6 AFAC2 Alberta FA benefit per child aged 7 11 AFAC3 Alberta FA benefit per child aged 12 15 AFAC4 Alberta FA benefit per child aged 16 17 Qu bec QFFSL Federal family allowance Quebec configuration QFS Federal supplement per child 12 17 on Quebec family allowance QFPSL Quebec provincial family allowance QAAFAFLAG Availability allowance supplement to Quebec FA flag QAAFA Quebec availability allowance Supplement QNBFAFLAG Quebec newborn family allowance flag QNBFA Quebec newborn family allowance Repayment through tax syste
14. Method of creating base variables VARMETH 2 3t Method of creating variant variables 4 INPVARMPR SPSD ba92 mpr Name of variant tax transfer Output Files OUTMRSFLAG 0 Variant results file creation flag OUTCPR factc1 cpr Name of control parameter file out Introduction and Overview Guide Page 59 SPSD M Version 9 2 5 6 7 8 9 10 11 10 11 12 Do you wish to modify any control parameters gt y By default variant results will be produced using the 1992 default tax transfer parameters found in the file spsd ba92 mpr These are the settings we shall use We shall now proceed to edit the control parameters for the FACTC1 run gt cprdesc Description of SPSM run string CPRDESC gt Base vs No Federal FA double CTCPC 5 sample basmeth Method of creating base variables BASMETH gt 2 inpbasmpr Name of base tax transfer parameter file in string INPBASMPR gt spsd ba92 mpr gt outmrsflag Variant results file creation flag OUTMRSFLAG gt 1 gt uvarflag Activate UVAR parameter for expressions UVARFLAG gt 1 gt uvar UVAR gt User statements string 01 deltacon immicons _immicons 02 label deltacon Change in Consumable Income deltacon immicons immicons label deltacon Change in Consumable Income The CPRDESC parameter is changed as usual The default setting for this parameter is 0 which causes no base results to be ei
15. The user should first create a separate directory for this test session by typing MKDIR user 1 If you already have a directory or a file with this name you will have to pick another name Next the user should make the new directory the current directory 2 1 C gt MKDIR user ENTER 2 C gt CD user ENTER 3 CAUSER spsm ENTER The SPSM is now invoked by typing spsm 3 A copyright message will be temporarily displayed followed by the SPSM start up screen as shown on the following page f this is not the case refer to the nstallation Guide for possible problems and solutions Welcome to the Statistics Canada Social Policy Simulation Database Model SPSD M Program Version 9 0 Database Adjustment Algorithm Standard Standard Tax Transfer Algorithm Version 9 0 88 05 Alternate Tax Transfer Algorithm none aucun Compiler Operating System MSC 12 0 WIN NT Copyright c Minister of Industry 1984 2001 4 Press F1 for context specific help 5 Press Ctrl Break to terminate LLLLLL lLllllll 2 l Control Parameters 6 Enter name of input control parameter file gt spsd ba92t Loading control parameters from spsd ba92t cpr 7 Enter specification for generating output files gt firstime Generating default output file names The information on the start up screen after the copyright can be safely ignored until the user makes glass box changes The h
16. also presented in the other guides in a more comprehensive fashion Addendum Introduction and Overview Guide Page 80 SPSD M Version 9 2 This document outlines the modifications that the SPSD M has undergone subsequent to the previous release of the SPSD M Users will note that the chapter headings in this Addendum correspond to guides in the manuals SPSD M USER S MANUAL The SPSD M has been designed and created by senior analysts who actively perform simulations using the tool It contains powerful features which allow a broad range of analysis options Comprehensive descriptions of these features are contained in the five guides in this manual User s Guide This guide describes how to run the SPSM in black box mode i e without making any changes to the C language source code The SPSM user facilities are each described in a complete fashion User X tab Facility This auxiliary guide contains an expanded more pedagogical treatment of the user defined X tab facility which is described in a complete though terse fashion in the SPSD M User s Guide Programmer s Guide Intended for use by the glass box user this guide describes how to make programming changes to the SPSM model Commodity Tax Model User s Guide This guide describes the operation of the Commodity Tax COMTAX model It also contains the mathematical description of the underlying Input Output model Tools User s Guide This guide describes various tools a
17. at the end of this guide as well as in the Variable Guide Level of Analysis The tax transfer system considers the family context of individuals in determining the value of many taxes and benefits For example child tax credits depend on the presence of children and the amount of family income The SPSM processes households and reports results based on several different social units The SPSD is a hierarchical family file in which the family relationships among all the members of a household are known The phrase a level of analysis refers to one of five hierarchically arranged types of families The individual is the smallest and a household is the largest The SPSM five family levels of analysis are defined as follows Individual Single person or record on the SPSD Nuclear family A head spouse if present and never married children under the age of 18 sharing the same dwelling Census Family A head spouse if present and never married children of any age sharing the same dwelling Economic Family A group of individuals living together who are all related by blood marriage or adoption and share the same dwelling Household Any individuals or group of individuals who share the same dwelling Note that unattached individuals are considered to be special kinds of census and economic families in SPSD M This usage differs from most Statistics Canada publications The SPSD data are stored at either the individual or the hou
18. been replaced by characters SPSM spsd ba92t test1 n n n n The second batch mode option is to create a file which contains the responses for a complete dialogue session using a full screen editor Using the example above such a file would look like this spsd ba92t testi n n n n If we had created a file with the above contents called TEST CMD we could then run Introduction and Overview Guide Page 34 SPSD M Version 9 2 the SPSM using this file by typing SPSM TEST CMD The use of batch mode and especially CMD files can assist greatly in organizing and documenting your work Batch mode lets you avoid repetitive dialogue sessions It also allows you to prepare a number of SPSM program runs which may then run unattended overnight on your PC SCREENING RECORDS A user will not always wish to simulate scenarios for every individual on the database either because of the time required or because the simulation is applicable to only some sub group of the population In these instances the user will want to make use of the sub sample facility or the selection facility By using random number stream variables the user may employ these facilities to create take up rates in black box mode Sub Sampling The SPSM can take a considerable amount of time to compute results for the entire SPSD database particularly if computational intensive facilities have been activated In order to quickly assess the validity of a run or to ex
19. deltadsp units L Mean Change Disposable Income totalfa deltadsp units L Effective Income Tax on FA deltacon deltadsp units L Effective Sales Tax on FA NF dispgrp deltadsp units L Mean Change Disposable Income iskid units P 2 L Avg children per family units L Census Families scfrecs L SCF Records Note that the hard wired table whose production is requested in the default control parameter file ba92t cpr is being turned off here No built in tables will be produced The flag to enable the production of user specified tables is turned on The table specification discussed next is fairly complex and readers already feeling a little overwhelmed may wish to proceed to the execution phase of the example on the next page This is the specification for the first of two custom tables It is given as one long string parameter XTSPEC which is continued onto subsequent lines The first table will be based on the nuclear family unit of analysis NF and will produce statistics broken into different rows by census family type with a total row at the bottom nftype The various columns which will be produced are shown on separate lines in session2 cpi to enhance readability For example the first column will display the average federal family allowances for census families with children The second table will produce results by level of disposable income The file session2 cpi described above can be fo
20. directory All these commands take as their argument a directory name For example MKDIR USERS will create a directory called USERS one level below the root The command CHDIR USERS will make USERS the current directory If you wish to display the name of the current directory simply enter the CHDIR command with no arguments The command RMDIR USERS would then delete the USERS directory Note that you cannot delete the root or the current directory The RMDIR command will not remove a subdirectory that contains files or other subdirectories thus users must delete any contained files before removing a subdirectory SPSM STRUCTURE As already mentioned the SPSD M models are configured as two distinct executable computer programs The central program SPSM calculates personal income taxes sales taxes and government cash transfers The other program COMTAX executes the macro economic commodity tax model that provides parameters for input to the SPSM The COMTAX program is not available on the demonstration disk or the full package and its execution is not discussed here However results from it are incorporated into parameter files in the form of effective retail sales tax equivalent rates Default SPSM files apply these parameters to household expenditures to estimate individual commodity taxes Any execution of the SPSM program has the following basic structure Introduction and Overview Guide Page 9 SPSD M Version 9 2 SPSD Databas
21. directory It is any directory other than the root The current directory is the directory in which you are currently working The current directory is the directory which MS DOS accesses when you specify a file without specifying a directory When you turn the computer on the root directory is the current directory until you change to another directory using the MS DOS CHDIR command A parent Introduction and Overview Guide Page 8 SPSD M Version 9 2 directory is the directory one level above the current directory in the tree structured hierarchy Pathnames Before MS DOS can find a file in a directory that is not in the current directory you must tell MS DOS where to locate it You do this by specifying a pathname A pathname consists of a drive specifier a series of directory names and a filename A backslash separates each of the elements For example the following pathname specifies that the file README TXT is located in disk drive D in the directory SOFTWARE which has the parent directory USERS The root is always indicated by a single backslash D USERS SOFTWARE README TXT Using Directories MS DOS uses several commands to list create locate change and delete directories The key actions are to create delete and change directories Command Short Form Purpose MKDIR MD Add a new directory to the tree structure RMDIR RD Remove a directory from the tree structure CHDIR CD Display the current directory or change the current
22. names 4 The SPSM start up screen provides information on the executable program The program version reflects the Statistics Canada release number 5 6 These three algorithm labels describe the set of C Language programs used to create the executable program These programs can be modified and recompiled in glass box mode 7 No glass box alternative algorithm exists in this version 8 The name of a valid control parameter file cpr must be specified with a Introduction and Overview Guide Page 44 SPSD M Version 9 2 path name if the file is not in the current directory The ba92t cpr file contains default settings for the status quo 1992 tax transfer system with sales taxes Note that the cpr extension can be omitted from the file name A valid MS DOS filename should be specified The default filenames are set to this filename with an extension appropriate to the type of output file Given the previous dialogue the SPSM then finds the specified control parameter file and displays some of its key contents as follows 10 11 12 13 10 11 12 13 Current values of selected control parameters CPRDESC Tax ransfer 1992 Population 1992 Incomes 1992 1992 5 LICENSEE Statistics Canada SPSD M licensee Selection Sample SELFLAG 0 Selection facility activation flag SAMPLEREQ _ 1 000000000 Size of sample requested INPSPD SPSD v90y97t spd Name of SPSD file in INPFXV SPSD v90y97t fxv
23. on the line labeled Change in Consumable Income in the table result file for the FACTC1 run This number must be brought to zero to obtain neutrality with respect to consumable income in the 1992 tax transfer system Dividing this difference by the rate calculated in step 4 yields an estimate of the amount by which CTCPC must be changed to get revenue neutrality Step 6 Perform a final run FACTC3 with the revised value of CTCPC calculated in Step 5 This run will only be approximately neutral with respect to total consumable income Examine the resulting tables FACTC3 TBL to determine the redistributive effects by province and income In summary three distinct SPSM runs will be performed The FACTC1 run makes a rough guess about CTCPC The FACTC2 run computes the effect on consumable income of changing CTCPC by 1 dollar The FACTC3 run uses a refined estimate of CTCPC using the derivative deduced from the FACTC2 run The following table summarizes the three runs Introduction and Overview Guide Page 58 SPSD M Version 9 2 Run Name Base Results file Variant Results file FACTC1 Standard 1992 ba92 mpr 1992 No FA Double CTC factc1 mpr FACTC2 1992 No FA Double CTC 1992 No FA Double CTC plus 10 factc1 mrs factc2 mpr FACTC3 Standard 1992 ba92 mpr Final CTC estimate factc3 mpr We will now proceed with the FACTC1 run 1 C SPSMTEST gt SPSM Social Policy Simulation Database Model SPSD M Program Version
24. read immediately after installing the SPSD M and before you attempt to use the program for the first time Most of the material contained in this guide is covered in greater detail in the User s Guide and Reference Manual What we do here is present you with an introduction sufficient to get you up and using the model quickly Once familiar with the SPSD M users of the full package can restrict their use to the User s Guide and the Reference Manual This software is complex The SPSD M is a highly flexible tool with a vast number of independently controllable interrelated parts It is designed to be easy to use With extensive use of default settings the flexibility is controlled by first time users However implicit in these settings are numerous assumptions which could potentially affect interpretation of the results Moreover the institutions and programs modeled by the SPSD M are themselves complex It is thus easy to make subtle mistakes which may result in significant estimation errors To make the best use of the SPSD M please read the available documentation Then if you have any problems or you are unsure of the simulation results please phone us at 613 951 3774 or email us at sosdm statcan ca WHAT YOU SHOULD ALREADY KNOW By now you should have either the demonstration or the full version of the SPSD M installed on your computer If you do not please do so now by following the instructions in the Installation Guide The g
25. to a reasonable value CASE OUTPUT FILES In addition to tabular data the SPSM can write out information from the database and model to a data file in a number of different formats These files contain the specified sub sample of an individual SPSM program run and can be used for subsequent analysis using either the SPSM a text editor or another software product The cases written can be at any level of analysis and may include any database or modeled variables For example a SAS file could be created containing the age sex census family type and incomes of all individuals aged 18 to 64 Alternately a file containing the urbanization class province residential tenure number of children and government cash transfers of all census families in Alberta could be produced With the exception of SPSM results files all files can contain user defined variables Users should be aware that the files created with this facility are not as compressed as are the main SPSD and SPSM result files For this reason they can easily exceed five megabytes and should be created with regard to the available space on your hard disk For example even a SAS file for all individuals with only three variables per individual can occupy 750K on disk Text Output Facility csv The SPSM text output facility creates a character file containing case by case micro data information One can either print or observe it directly or one can use it as input into some other c
26. which processes each individual and family on the SPSD calculates taxes and transfers using algorithms that simulate legislated or proposed programs and reports on the results A sophisticated software environment gives the user a high degree of control over the inputs and outputs to the model and allows the user to modify existing programs or to examine proposals for entirely new programs Within the SPSM there are two models configured as two distinct computer programs e The central program the SPSM is a microsimulation model which calculates taxes and transfers for individuals and families These calculations are performed for everyone on the SPSD and then aggregated to obtain estimates The SPSM is a static impact model and is therefore not intended to simulate how an individual s behavior is likely to change in response to various policy options The SPSM program includes software which allows the user to perform tabulation and database retrieval operations on the database information e The commodity tax model COMTAX is a macro economic Input Output based model This model is not currently released with the SPSD M package but rather the results from it are COMTAX provides estimates of effective federal and provincial retail sales tax equivalent commodity tax rates by province household expenditure category and tax type This model is required because many commodity taxes are levied at intermediate stages of production not at the fi
27. while no underscore refers to the variant variable value deltadsp is the corresponding variable for disposable income 10 We will report on both the federal family allowance program as well as all provincial Quebec family allowance programs Since no combined variable exists it is created here 11 We wish to generate statistics on the average number of children in a census family However the variable cfnkids number of children in the census family is a classification variable and cannot have mathematical actions performed on it Therefore we create a new analysis variable iskid which is simply a flag which is set to one if the individual is less than age 18 and is zero if the individual is age 18 or over This variable may then be added up etc in the X tab facility 12 We will want to report on losses of family allowances by level of nuclear family disposable income Because disposable income is an analysis variable we must specify the desired breakpoints in this continuous variable The breakpoints are arbitrary and use 5 000 intervals Introduction and Overview Guide Page 50 SPSD M Version 9 2 13 14 15 16 13 14 15 16 2 1 13 Reports 2 1 15 1 Built in tables TOFLAG 0 Table 0 request flag 3HE 2 1 15 2 User specified Tabulation facility XTFLAG 1 X tab facility activation flag XTSPEC NF nftype totalfa units L Base Average FA deltacon units L Mean Change Consumable Income
28. 1 We shall do the analysis using the 1992 tax transfer system and 1992 demography This eliminates any need to project the values of or grow money items The overall results can be used as the basis for estimating the impacts in other years Alternatively growth factors tax transfer parameters and weights for the desired target year could be used in a subsequent study to refine the overall results Step 2 Perform an SPSM program run with base results for the 1992 system and variant results with federal Family Allowances eliminated and the Child Tax Credit amount CTCPC doubled Create the variant results file FACTC1 MRS We shall call this run FACTC1 and run it using the 5 sample Step 3 Perform a second SPSM program run similar to the FACTC1 run except that the base results are read from the results file FACTC1 MRS and the variant results are based on the same parameter set except that CTCPC is increased by a further 10 dollars We will call this run FACTC2 Step 4 Examine the total amount on the line labeled Change in Consumable Income in the table output file for the FACTC2 run This number will be directly attributable to the 10 dollar increase in CTCPC since that was the only tax transfer parameter that was changed The difference divided by 10 i e the derivative is the amount by which total consumable income measured in millions of dollars will increase if we increase CTCPC by 1 dollar Step 5 Examine the total amount
29. 5 36 50 2001 SPSM completed on Fri Sep 14 15 36 50 2001 Elapsed time 00 00 02 Number of households processed 3990 Number of persons selected 11954 Variant consumable income 000 000 364234 0 Control parameter file test1 cpr Table output file test1 tbl Variant tax transfer parameter file test1 mpr Variant results file test1 mrs C SPSMTEST gt dir Volume in drive C has no label Volume Serial Number is F884 44FE Directory of CAspsmtest 09 14 2001 03 36p lt DIR gt 09 14 2001 03 36p lt DIR gt 09 14 2001 03 36p 12 432 test1 cpr 09 14 2001 03 36p 225 706 test1 mpr 09 14 2001 03 36p 119 759 test1 mrs 09 14 2001 03 36p 10 677 test1 tbl The SPSM produces an estimated completion time which usually appears one minute after the start of simulator execution This is not present here because of the speed of the computer which ran the demonstration The SPSM also reports on the simulated income in millions of dollars If a base and variant are run the model also reports on the difference in consumable income between the two scenarios The SPSM always indicates the files that have been output These files have all been written to the current directory as can be seen by checking the SPSMTEST directory s contents In this example the MS DOS dir command is invoked after the SPSM has finished to display the contents of the SPSMTEST directory The table file will be identical to that created in the quick start example in Sectio
30. 50 0 95650 Introduction and Overview Guide Page 53 SPSD M Version 9 2 22 23 24 25 21 22 23 24 25 26 Do you wish to modify any database adjustment parameters gt n seher Tax Transfer Parameters Loading variant tax transfer parameters from SPSD ba92 mpr Current values of selected variant tax transfer parameters MPRDESC Current values for 92 CTFLAG 1 3t Commodity tax activation flag CTOPT 2 Commodity tax calculation method CTDFLAG 0 Commodity tax detailed calculation flag TARGETYEAR 1992 Year of analysis Do you wish to modify any variant tax transfer parameters gt y mprdesc Description of tax transfer parameter file Source Given as LABEL 3t Update Value LABEL MPRDESC gt 1992 No Family Allowances gt faflag Family allowance flag Source To calculate Federal and Provincial FA FAFLAG gt 0 gt go Writing out changed tax transfer parameters to test2 mpr Changing INPVARMPR from SPSD ba92 mpr to test2 mpr The default parameter settings for the 1992 database adjustment parameters will cause database adjustment to be performed The actual parameters are mostly set to 1 Some expenditure adjustment parameters have however been set to scale up total expenditures for known underreporting problems Default database adjustment settings will be used The variant parameters must be modified in order to simulate the elimina
31. 9 To proceed to the next phase of the SPSM dialogue the command go is entered 20 The user is prompted to be sure that no further control parameter changes are needed The usual response is no or n The program then writes out a copy of the modified cpr parameter file and proceeds to the next phase of the dialogue Introduction and Overview Guide Page 46 SPSD M Version 9 2 Database Adjustment Parameters 21 Loading database adjustment parameters from SPSD ba97_92 apr Current values of selected database adjustment parameters GFISENF 10 Growth Factor Self employed income non 0 74940 0 89140 0 85640 0 79350 0 87880 0 77390 0 79200 0 75870 0 79820 0 85950 GFIEMP 10 Growth Factor Employment income 0 96450 0 96450 0 99370 0 99370 0 99370 0 99370 0 96230 0 96230 0 95560 0 95560 0 94580 0 94580 0 97530 0 97530 0 88710 0 88710 0 90040 0 90040 0 95650 0 95650 22 Do you wish to modify any database adjustment parameters gt n BEBO Tax Transfer Parameters Loading variant tax transfer parameters from SPSD ba92 mpr Current values of selected variant tax transfer parameters MPRDESC Current values for 92 23 CTFLAG 1 Commodity tax activation flag CTOPT 2 Commodity tax calculation method CTDFLAG 0 Commodity tax detailed calculation flag TARGETYEAR 1992 Year of analysis 24 Do you wish to modify any variant tax tran
32. ATING SYSTEM CONSIDERATIONS All SPSD M files are maintained using a directory structure that follows certain conventions Under Windows NT 4 0 and Windows 95 98 directories are referred to as folders Users are strongly urged to organize their work in directories as well Directories provide a structure for organizing files on disk If you are unfamiliar with the hierarchical file system the figure on the following page provides a brief introduction If you are still unclear please refer to the operating manual which came with your computer Executable files have either a exe com or bat extension They are executed by simply typing the file name without the extension followed by a carriage return at the MS DOS prompt C gt if your default disk drive is your hard disk which it should be The spsm exe program is an example of such a file The full pathname of the executable SPSM program is spsm win32 spsm exe During the user dialogue several specifications of input and output files are required If the user does not specify a full pathname when specifying a file the file will be read from or written to the current directory The following figure shows the SPSD M Demonstration package directory structure and selected file contents together with a user created directory ROOT SPSM SPSD USER SPSME HLP BA97T CPR FIRSTIME CPR SPSMF HLP BA97 APR FIRSTIME TBL BA97 MPR V90Y97T SPD V90Y97T FMX d B V90Y97T WGT SP
33. D file in INPFXV SPSD v90y97t fxv Name of FAMEX vector file in INPWGT SPSD v90y92t wgt Name of weight file in Growth INPAPR SPSD ba97 92 apr Name of database adjustment parameter file Scenario Control no base variant computed using parms gt BASMETH 0 Method of creating base variables VARMETH 2 Method of creating variant variables INPVARMPR SPSD ba92 mpr Name of variant tax transfer parameter file Output Files OUTMRSFLAG 0 Variant results file creation flag OUTCPR firstime cpr Name of control parameter file out For now this information can be ignored as we are not interested in changing any of the default settings However users are reminded that we are requesting a 10096 sample of the SPSD M 596 sample database 8 9 Do you wish to modify any control parameters gt n No files will be overwritten with these control parameters 10 Any further control parameter changes gt n Writing out control parameters to firstime cpr To use the default files or files prepared previously with a text editor no changes are required using the dialogue manager 9 A no response to the next prompt 10 then causes the dialogue to proceed to the next phase Introduction and Overview Guide Page 13 SPSD M Version 9 2 E S S S ZZ Database Adjustment Parameters 11 gading database adjustment parameters from SPSD ba97 92 apr gt Current values of selected database adjustment parame
34. ITY The distributional analysis facility calculates a set of univariate distributional statistics for a sample of SPSD cases If the distributional facility is activated a user specified random sample of up to 10 000 cases is stored and sorted in memory and distributional statistics are calculated on this sample The statistics provide the following information e Number of zero and non zero observations e Descriptive Statistics sum of weights sums sum of squares weighted mean e Extreme values bottom five values and top five values e Selected quantiles quintiles quartiles etc e Histogram plot All output from the distributional facility is written to the TBL file for the program run An example using the distributional facility is shown in tutorial Session 4 The following is an example of output from the distributional facility that has been produced for the variable imccea Child Care Expenses Allowed Introduction and Overview Guide Page 38 SPSD M Version 9 2 SPSD M Version 6 00 Base Description No base results Variant Description Current values for 1992 Sample 1 0000 AGENAME Standard adjustment Distribution report Child care expenses allowed for Individuals Total observations 200264 Zero observations 194998 The following statistics are based on 5266 non zero observations Descriptive Statistics Done for all records Done for Sum of weights 699640 all records Weighted Sum 1700095111 Weighted Sum o
35. MPLEREQ _ 1 000000000 Size of sample requested Input Databases INPSPD SPSD v90y97t spd Name of SPSD file in INPFXV SPSD v90y97t fxv Name of FAMEX vector file in INPWGT SPSD v90y92t wgt Name of weight file in Growth INPAPR SPSD ba97 92 apr Name of database adjustment parameter file Scenario Control no base variant computed using parms gt BASMETH 0 3t Method of creating base variables VARMETH 2 3t Method of creating variant Variables INPVARMPR SPSD ba92 mpr Name of variant tax transfer Output Files OUTMRSFLAG 0 Variant results file creation flag OUTCPR test2 cpr Name of control parameter file out Do you wish to modify any control parameters gt y From the selected list of parameters we can see that they are the same as the Session 1 example with the exception that test2 is specified for naming output files Introduction and Overview Guide Page 52 SPSD M Version 9 2 18 gt read Name of file to read gt spsm example session2 cpi parameter CPRDESC updated from file parameter VARDESC updated from file parameter BASMETH updated from file parameter BASDESC updated from file parameter INPBASMRS updated from file parameter SELFLAG updated from file parameter SELUNIT updated from file parameter SELSPEC updated from file parameter UVARFLAG updated from file parameter UVAR updated from file parameter TOFLAG updated from file parameter XTFLAG updated from file parameter XTSPEC updated from fil
36. SM WIN32 SPSM EXAMPLE SPSM EXE EG3DIST CPI SESSION2 CPI EG3XTAB CPI QFFSL MPI EITC CPI Figure 1 SPSD M Directory Structure and Selected File Contents When about to begin an analysis using the SPSM the user should create a new directory to store all the input parameter files and outputs Any name may be Introduction and Overview Guide Page 7 SPSD M Version 9 2 selected for this directory In Figure 1 the user directory is assumed to have been created for this purpose using the MS DOS mkdir command Directories containing large numbers of files are difficult to organize and impair MS DOS efficiency A further level of sub directories is recommended if the analysis requires many files An Overview of Selected Features of MS DOS MS DOS is an operating system that controls your computer and the devices connected to it When you use your computer you are using and or creating files of related data These files are then managed using MS DOS Files A file is a collection of stored information Because you can store many files on disk naming files is an important task Good filenames are descriptive and are composed of between one and eight characters Spaces may not be used in filenames check your MS DOS manual if you are unsure of the valid characters A filename may also have an extension although one is not required An extension always begins with a period followed by a maximum of three characters The extension allows you to gr
37. SM completed on Mon Sep 17 08 34 54 2001 Elapsed time 00 00 01 Number of households processed 3990 Number of persons selected 11954 Variant consumable income 000 000 364234 0 Control parameter file firstime cpr Table output file firstime tbl The default tax transfer parameters are read as specified in the control parameter file Sales taxes will be calculated 13 as the flag for commodity taxes has been set to 1 by default Because we wish to examine results using the 1992 tax transfer system no parameter changes will be made 14 Control is then passed back to SPSM for completion of the simulation Introduction and Overview Guide Page 14 SPSD M Version 9 2 The numbers seen below the SPSM started row represent the estimated percent of the run completed and are covered over with symbols as the run progresses 15 If at any time the user wishes to stop or pause execution and examine some key results as calculated so far he or she should press CTRL BREAK press the BREAK key while holding down the CTRL key The program will pause its computations display current totals and prompt the user whether to continue or not If the user decides to stop partway through the execution output files are immediately written and control parameter files are altered to reflect the portion of the database actually read The program always ends with some key information on the run including a list of all output files
38. SPSM can be run in batch mode to allow a series of simulations to run unattended We have tried to make the documentation as clear and concise as possible and the software has been extensively tested Still errors are possible If you encounter any problems with the software or documentation contact Statistics Canada directly at 613 951 3774 Quick Start The SPSM can be very simple to operate and can give fast results A basic understanding of the MS DOS operating environment and the SPSM structure is all that is required It should be kept in mind however that the SPSM is also a highly complex tool The multi level database and numerous parameters and facilities give great flexibility but can lead to confusion if many simultaneous changes are attempted For this reason the SPSM makes extensive use of default parameter files and built in documentation features The user can always begin from the 1997 defaults alter various settings and examine the results The supplied SPSD M 1997 default parameter files are configured to simulate the 1997 legislated tax transfer system while producing one default table After first briefly examining operating system considerations and SPSM structure this section will describe how to make a program run using these default files The section concludes with an examination of the output and a general discussion of how to view your results Introduction and Overview Guide Page 6 SPSD M Version 9 2 OPER
39. Scenario Control no base variant computed using parms BASMETH 2 3t Method of creating base variables INPBASMPR SPSD ba92 mpr 4 Name of base tax transfer parameter file VARMETH 2 3t Method of creating variant variables INPVARMPR factc1 mpr Name of variant tax transfer Output Files OUTMRSFLAG 1 Variant results file creation flag OUTMRSVARS _ immicons Variant results file variables OUTVARMRS factc2 mrs Name of variant results file out OUTCPR factc2 cpr Name of control parameter file out Do you wish to modify any control parameters gt y The values of the selected parameters are the same as for the FACTC1 run as expected The output files are correctly named for this run gt cprdesc Description of SPSM run string CPRDESC gt No Federal FA Double CTC vs same with CTC 10 basmeth Method of creating base variables BASMETH gt 1 gt inpbasmrs Name of base results file in string INPBASMRS gt factc1 mrs gt inpvarmpr Name of variant tax transfer parameter file in INPVARMPR gt factc1 mpr gt outmrsflag Variant results file creation flag OUTMRSFLAG gt 0 gt go No files will be overwritten with these control parameters Any further control parameter changes gt n Writing out control parameters to factc2 cpr The CPRDESC parameter is changed as usual Base results were calculated in the FACTC1 run In this FACTC2 run results will be read from a results file In
40. The increase in consumable income shown here will be slightly different than the corresponding figure in the output table file This disagreement is smaller than rounding error and is caused by slightly different means of applying weights to the estimates The following is a listing of the first table in the FACTC2 TBL file created by the SPSM run SPSD M Version 9 00 Base Description ba92t No Federal FA CTC 2 Variant Description No federal FA Double CTC plus 10 Sample 0 0510 AGENAME Standard Table 1U Selected Quantities for Census Families dM TM Base Consumable Income M 365958 4 Variant Consumable Income M 366016 8 Change in Consumable Income M 58 3 oommMMMmMMMMMee e The table reveals that consumable income increased by 58 3 million over the FACTC1 run Since the CTCPC parameter was increased by 10 this means that each 1 increase in CTCPC results in an increase of 5 83 million in consumable income Thus to eliminate the 1724 4 million overshoot of the FACTC1 run we should decrease CTCPC by 1724 4 5 83 295 78 dollars Since the FACTC1 run had CTCPC set at 1202 our new estimate for CTCPC for the next run FACTC3 should be 906 25 1202 less 295 78 Introduction and Overview Guide Page 66 SPSD M Version 9 2 The following is a listing of the include file spsm example eg3xtab cpi which will be read by SPSM during the FACTC3 run HHHH eg3xtab cpi Include file for Tutori
41. al Session 3 TH SId TH Purpose This file contains the necessary control parameters to TH create user defined variables and create a custom table THE using the X tab facility The explanation of the HE parameters may be found in the Introduction and Overview TH guide under tutorial session number 3 PELE TEE A TEA EE A a a HAHRRRRHAHH HAHHA HRHHHHEH UVARFLAG 1 UVAR Deltacon immicons immicons label deltacon Change in Consumable Income deltafb imfedbal imfedbal label deltafo Federal Net Balance Increase deltapb imprvbal imprvbal label deltapb Provincial Net Balance Increase iconsgrp split _immicons 5000 10000 15000 20000 25000 30000 35000 40000 45000 label iconsgrp Base Consumable Income Group XTFLAG 1 XTSPEC CF hdprov deltafb deltapb deltacon CF iconsgrp deltacon L Change in Consumable Income This include cpi file is designed to produce results which will allow us to assess the impact of our changes on the progressivity of income distribution and on federal and provincial treasuries The three user defined variables correspond to each of these items of interest The simple X tab specification is all we will need to perform the final analysis We can now proceed with the final SPSM run in this example The program is invoked as usual CASPSMTEST SPSM Welcome to the Statistics Canada Social Policy Simulation Database Model SPSD M Program Version
42. alculate Base Scenario Calculate Variant Scenario Calculate User Variables Selection and Output Routines Variant Database Base Variables User Reference Results Results Variables Variables ie Variables Available to output routines gt T1Tax UI Family Data Histories Expenses PARAMETERS Parameters are used by the SPSM to control every aspect of a program run Introduction and Overview Guide Page 25 SPSD M Version 9 2 Input Output Tables Tax Transfer Parameters Scenario Database Control Adjustment Parameters Parameters Income Payroll Commodity Taxes amp Cash Taxes Transfers SNA Aggregates Tax Forms Conference Budgets Board Forecasts e Redbook SPSM Parameter Types As mentioned earlier the SPSM is driven by over 800 parameters divided into three groups corresponding to three types of files that control three main processes Control parameters specify input and output files for a specific model run and are used to activate model software facilities Database adjustment parameters control the inflation of database money items Tax transfer parameters control the specific functioning of the tax and transfer programs In the case of commodity taxes a set of default parameters is supplied Perusing the included lists of the tax transfer parameters at the back of this guide can give one a feel for the scope of options available i
43. and output files These conventions have been developed based on a great deal of experience with the SPSM While SPSD M files names and extensions are not restricted in any way the user is strongly urged to become familiar with these conventions before attempting to design some alternate system Database Inputs SPD SPSD binary database file ane FAMEX household level expenditure data SPSD file containing survey weights Parameter Inputs Complete and mandatory CPR Control parameter file APR D MPR atabase adjustment parameter File Tax transfer parameter file Partial and optional CPI Control parameter include file API D MPI atabase adjustment parameter include file Tax transfer parameter include file SPSM Outputs Printable TBL Output tables result file PRN Case output facility text results file Input to further analysis MRS Binary SPSM results file ae SAS fil SFM IS SAS formats source code output file Program Run Control CMD SPSM Dialogue commands Figure 7 SPSD M File Naming Conventions The database files are all named by Statistics Canada and should not be changed The parameter files with CPR APR and MPR must be complete and valid SPSM parameter files and values for all three must be specified for any given program run The CPI API and MPI files are created with a text editor and contain subsets of the corresponding CPR APR and MPR files which may be read in during the user dialogue Consult t
44. be used in the parameter files themselves because the SPSM will strip these block comments lines starting with from the output parameter file Introduction and Overview Guide Page 74 SPSD M Version 9 2 3 4 5 6 7 8 9 The selection facility will be used to perform a portion of the eligibility test The selection level of analysis is set to census families SELUNIT 2 The selection specification checks for individuals under the age of twenty one whose relationship to the head of the census family is that of a child As a result only those census families with children under 21 will be selected More precisely if SELSPEC evaluates to zero for every member of a specific SELUNIT a census family in this case then that family will be rejected THHE 2 1 9 User defined Analysis Variables THE UVARFLAG 1 UVAR nfempinc idiemp SP idiemp if idage gt 21 if nfempinc lt 8000 eitc 15 nfempinc else if nfempinc lt 12000 eitc 1200 else if nfempinc lt 24000 eitc 1200 nfempinc 12000 10 else eitc 0 label eitc Earned income tax credit The benefits are added up into one variable for reporting purposes 3 The user defined variable facility is used to simulate the EITC program Keep in mind that each user defined variable is evaluated sequentially at the individual level of analysis Consequently expressions should not make reference to user variables
45. dard Standard Tax Transfer Algorithm Version 9 0 88 05 Alternate Tax Transfer Algorithm none aucun Compiler Operating System MSC 12 0 WIN NT Copyright c Minister of Industry 1984 2001 Press F1 for context specific help Press Ctrl Break to terminate Introduction and Overview Guide Page 63 SPSD M Version 9 2 21 21 22 22 23 24 25 26 27 28 23 24 Enter name of input control parameter file gt factc1 cpr Loading control parameters from factc1 cpr Enter specification for generating output files gt factc2 Generating default output file names We will start with the control parameter settings established in the FACTC1 run Many of the settings will be the same for this run and we shall save time editing the parameters For example we shall be using the same user defined variable and X tab specification Output files will be based on the name FACTC2 Current values of selected control parameters CPRDESC Base vs No Federal FA double CTCPC 5 sample LICENSEE Statistics Canada SPSD M licensee Selection Sample SELFLAG 0 Selection facility activation flag SAMPLEREQ 1 000000000 3t Size of sample requested Input Databases INPSPD SPSD v90y97t spd Name of SPSD file in INPFXV SPSD v90y97t fxv Name of FAMEX vector file in INPWGT SPSD v90y92t wgt Name of weight file in Growth INPAPR SPSD ba97 92 apr Name of database adjustment Parameter file
46. djustment oce sas dre E eevee is ees 22 Calculate Base Results and Variant Results 22 User Variables CE 24 Screening Cases and Output Phase ccceeeeeeeeeeeeeteeeeeeeeeeeeeees 24 ERAN ainera cree acct E 25 GonuolPAramslersd uto ee EE SAN Eque 27 Database Adjustment Parameters sssssssss 27 Tax Transter Paramelefs eiui a xe EV a d e ob eiie etus 28 evolat CELO 28 Disposable and Consumable Income essssssseeeeeeeeee 29 SPORIS vate duas dA Md E TR REM Ed E 29 SPSM User FaclliliBS onore ES roD DEDE DEI D Ime 31 User defined Variables eere eb e sepe sta epe restet eadds 31 Parameter Edilintg oin taeit tete ut edd dedecus ecd tat ua desides tectus 32 Screening RECOTd Sie ECC 35 User Defined Variable Facility cccccccceeeeeeeeeeeeeeeeceeeeeeeeeeeeesesneaeeeeeeees 36 Greating Tables ice eter Sees cee DD EPOD ERU DIDI p 36 Distributional Analysis Facility ccccseecceeeeeeeeeeeeeeeneeeeeeeeeeeeeeseenaeeeeeeees 38 Marginal Tax Rate Facility 39 Gase Qutp t Files e lceeie end aaea e 40 aidalla Bl rte gd eral EA EE E EE T 42 Goal Seeking Facility dob Sedet nette tete A A E 42 Reference Variable Facility cc ccccceeeeeeeeeeeeeeceeeeeeeeeeeseeeaeeeeeeeeeeeeeeees 43 Example IMU AMON 5 9 0 a dea dete cbr iod LE a Mad 43 Session 1 Create Variant Results File eeseeseesus
47. e 19 gt go 20 No files will be overwritten with these control parameters Any further control parameter changes gt n Writing out control parameters to test2 cpr 18 Now the session2 cpi file can be read The first step is to specify the read command You will then be prompted for the name of the file to read The dialogue manager then opens and reads the cpi file and updates the values of the 13 parameters we specified earlier and reports which parameters have been updated as shown on the next page 19 All our control parameter changes were contained in the session2 cpi file so we can proceed to the next phase of the dialogue 20 This message indicates that there are no existing files in the current directory which will be overwritten If files already exist which would be overwritten by SPSM the user is informed of this condition and is given the chance to change the output file names if desired MEE Database Adjustment Parameters 21 Loading database adjustment parameters from SPSD ba97_92 apr Current values of selected database adjustment parameters GFISENF 10 3t Growth Factor Self employed income non 0 74940 0 89140 0 85640 0 79350 0 87880 0 77390 0 79200 0 75870 0 79820 0 85950 GFIEMP 10 Growth Factor Employment income 0 96450 0 96450 0 99370 0 99370 0 99370 0 99370 0 96230 0 96230 0 95560 0 95560 0 94580 0 94580 0 97530 0 97530 0 88710 0 88710 0 90040 0 90040 0 956
48. e Table Outputs Microdata Outputs SPSM Parameters SPSM Results Figure 2 SPSM Program Run Structure Microdata on individuals and households as well as parameters are specified inputs to the SPSM Optionally results from a previous SPSM run may also be input into subsequent runs The SPSM then performs the simulation calculates results and outputs reports and data files The SPSM is driven by over 800 parameters that control three main processes Control parameters specify input and output files for a specific model run and are used to activate model software facilities Database adjustment parameters control the growth of database money items Tax transfer parameters control the specific functioning of the tax and transfer programs The SPSM creates a set of output files that the user has specified Some of these files may then be printed or examined with a full screen editor while others contain microdata for further analysis using the SPSM or some other software package Any SPSM program run consists of the following four basic steps Introduction and Overview Guide Page 10 SPSD M Version 9 2 1 Invoke SPSM 2 User Dialogue 3 Simulation 4 Generate Outputs Brief descriptions of these steps are given below 1 The SPSM is contained in an executable file called sosm exe in the spsm win32 directory The SPSM is started or invoked by typing SPSM followed by a carriage return enter in response
49. e a number of additional keys become active UP ARROW DOWN ARROW go to previous or next line in the parameter CTRL X Split line at cursor PgUp PgDn first and last line of parameter Help Facility At any time during the user dialogue the help facility can be accessed by pressing the F1 key The help facility provides context specific explanations of the various prompts encountered during the dialogue This includes information on the types of parameters the editing keys and editing procedures and the available commands A future release of SPSD M will include help information on all parameters and variables Batch Mode The dialogue manager is the only means by which input parameters are specified to the SPSM However there are three ways to provide information required by the dialogue manager one interactive and two in batch mode The first batch mode option is to include a command string as an argument when invoking the SPSM The command string will take the form of a series of responses to the prompts in the order in which they are required These are actually the exact keystroke sequences which would be used in an interactive session The following example will run the default simulation on the ba92t cpr control parameter file the same program run as was performed in an earlier example Note that the following is the exact sequence of keystrokes that the user typed during that user dialogue session except that carriage returns have
50. e Description Current values for 1992 Variant Description ba92t No Federal FA CTC 2 Sample 0 0510 AGENAME Standard Table 1U Selected Quantities for Census Families ToMMMMMMMM M MMMMMMM TM t Base Consumable Income M 364234 0 Variant Consumable Income M 365958 4 Change in Consumable Income M 1724 4 TM Examine the total amount on the line labeled Change in Consumable Income above This number must be brought to zero to obtain neutrality with respect to consumable income in the 1992 tax transfer system Here we see that our changes have resulted in a 1724 4 million dollar increase in consumable income the increase in the child tax credit was somewhat excessive Obviously the number will have to be adjusted downward by some amount This completes Step 2 of the analysis plan previously outlined Part 3 of the plan and the second SPSM run involves an increase in CTCPC by 10 to work out the relationship between an increase in CTCPC and the corresponding increase in consumable income Both the base and variant scenarios in this run will use the 1992 tax transfer system with no federal family allowances The only difference between them will be the value of the child tax credit per child At this point you should invoke the SPSM as usual CASPSMTEST gt SPSM Welcome to the Statistics Canada Social Policy Simulation Database Model SPSD M Program Version 9 0 Database Adjustment Algorithm Stan
51. e basis of parameter inputs alternate algorithms glass box or both Additionally through the use of Introduction and Overview Guide Page 5 SPSD M Version 9 2 auxiliary input databases users may compare the results of any number of different simulation scenarios in a single execution of the SPSM e A comprehensive set of default tables includes dollar amount and persons reporting for over 40 variables with breakdowns by province income class family type and proportion above or below specified income levels by family size and type e A cross tabulation facility allows the user to specify custom N dimensional tables using database model or user defined variables The user is given full control over output formats precision level of analysis and labeling of the tables e Output interface facilities allow the user to create custom files in compressed binary format any of three ASCII file formats or as a native SAS file These files are intended for subsequent analysis using the SPSM or other software packages and can contain any combination of database or modeled variables e Performance has been given high priority in the design of the SPSM but still varies greatly depending on the hardware environment the simulation request and the types of output specified The SPSD M modeling environment requires a Windows 95 98 NT compatible computer A full model run can be performed in under thirty seconds with a Pentium III system The
52. e income The second column represents the output value corresponding to the input value in the same column The third column represents the marginal change in the output value for the next increment or tax bracket TABLE A table is a numeric array with an arbitrary number of rows and columns The array is preceded by a number indicating the number of rows which follow Parameter editing mode is entered by providing a parameter name in response to the gt prompt There are two distinct editing modes In single line mode the parameter being modified is either a number or a short string Multi line mode on the other hand is used to view or modify a string parameter whose value may spill over into multiple lines Multi line mode can be recognized by a prompt of the form 1 The number in square brackets gives the current line of the parameter being displayed or edited When single line mode is first entered pressing a key for any printable character will erase the currently displayed value allowing the user to give a new value Alternatively a number of editing keys allow the modification of existing values These keys are as follows ESC discards changes to the line and restarts with the original value Introduction and Overview Guide Page 33 SPSD M Version 9 2 INS toggle between insert and overtype mode HOME END BKSP DEL RT ARROW LEFT ARROW have normal functions ENTER indicates that changes are complete In multi line mod
53. e their own estimates if they wish to supply alternate growth rates Tax Transfer Parameters The final set of parameters is for controlling the algorithms which simulate taxes paid to and cash transfers received from government Users provide values for the appropriate year of analysis or the particular tax transfer scenario of interest These parameters can switch entire programs on or off as well as specify benefit levels reduction rates turning points tax tables and so forth If for example a user wished to examine the impact of increasing the tax back rate for the child tax credit from five to ten percent and the benefit level by 10 he or she would change the values of the two tax transfer parameters CTCRR and CTCPC OUTPUT FACILITY The SPSM writes all requested simulation results to output files There are three main purposes for which the SPSM generates outputs e To document parameter settings which produce a program run e To provide case by case information on simulation results e To provide tabular aggregate information on simulation results Outputs which document the program run are stored in the various specified parameter files These parameter files are output in a self documenting format with parameters labelled and arranged in sections These files may be viewed on the screen or printed if desired An output control parameter file is always produced It contains among other things input and output files used in the ru
54. elp facility is easily accessed 4 and the user can break out of the program 5 at any time In the first phase of the dialogue the user creates or modifies a control parameter file The user must provide a valid control parameter filename in response to the first prompt In this example the default control parameter file for the 1992 tax transfer system is specified with its full pathname 6 Note that the default file extension of cpr can be omitted if desired The fact that we are requesting ba92t as opposed to ba92 indicates that we want the 5 sample database as opposed to the full database The second prompt requests a name to use when creating filenames for output from this run 7 If the user does not specify a response to this prompt SPSM will create output file names based on the full pathname of the input filename specified in 6 In this example we will use the name firstime Introduction and Overview Guide Page 12 SPSD M Version 9 2 After the user has responded to these two prompts the SPSM finds the specified control parameter file and displays some of its key contents as follows 8 Current values of selected control parameters Descriptive CPRDESC Tax transfer 1992 Population 1992 Incomes 1992 1992 5 LICENSEE Statistics Canada SPSD M licensee Selection Sample SELFLAG 0 Selection facility activation flag SAMPLEREQ 1 000000000 Size of sample requested Input Databases INPSPD SPSD v90y97t spd Name of SPS
55. er Variables Although there are hundreds of variables available in the SPSD M users will often wish to create their own variables for reporting purposes This is done by manipulating the database and modeled variables The user may create classification variables by specifying a continuous or discrete variable together with a set of breakpoints which define the class boundaries Analysis variables may also be created by specifying an expression which manipulates logically and or algebraically any of the database or modeled variables User variables may be given names and labels as well as level labels for classification variables Screening Cases and Output Phase Users may wish to restrict the calculation of results to a certain subset of households or to a certain type of individual or family on the database This can serve both efficiency and analytical purposes This screening can be based on modeled as well as database variables Therefore it follows the calculation of base and variant results For example households having no children may be selected or individuals with no employment income excluded Screening affects all of the SPSM output facilities in any given program run The output phase is the last step of the simulator and is designed to produce a series of outputs for different purposes Output is either intended to be looked at or Introduction and Overview Guide Page 24 SPSD M Version 9 2 printed on the one hand or to be input
56. erleaved with the dialogue The notes make general comments concerning the SPSM as well as pointing out specific features highlighted in the tutorial example The symbol CTRL X is used to indicate that the user has typed the X key while holding down the CTRL key ENTER indicates that the user has pressed the enter key Before you begin you must have correctly installed the tutorial version of the SPSM with the INSTALL program following the instructions in the nstallation Guide Ensure that MS DOS environment variables and configuration information have been set correctly The text and results obtained with your computer may not exactly match the results printed in this guide The results on your screen if the example steps are followed exactly should be interpreted as correct This is due to the fact that the documentation sometimes lags the most recent version of the software Please call Statistics Canada at 613 951 3774 should you have any questions SESSION 1 CREATE VARIANT RESULTS FILE This example makes a simple single program run with the SPSM We will establish a new working directory and create an SPSM results file a control parameter file Introduction and Overview Guide Page 43 SPSD M Version 9 2 and an algorithm control file and a table file This example highlights the dialogue manager and the steps required for a normal program run The first step in our example includes creating a new working directory and invokin
57. f Squares 642937195486 Weighted Mean 2430 Extreme Values with associated household numbers Minima hdseqhh Maxima hdseqhh 8 17244 12000 7376 8 17245 12000 17870 8 59930 12000 20694 33 22107 12000 27402 44 59185 12000 27403 Done for The following statistics are based on a sample of 3000 sample Selected Quantiles Q1 986 P1 320 P90 4574 P20 858 P60 2294 Med 1833 P2 367 P95 6000 P30 1144 P70 3153 Q3 23520 P5 440 P98 8000 P40 1467 P80 3740 P10 587 P99 8000 Note that the maximum value on the database is equal to a full deduction for three children If extreme values are not as anticipated the user can then produce a detailed report on the relevant households by using the selection facility and the print file output facility Histogram Plot eee pie prope P1 P20 P40 P60 P70 P80 P90 P95 P99 320 858 1467 2294 31533740 4574 6000 8000 The distribution report has been produced for all non zero values of imccea showing the skewed truncated shape of the density function for the tax expenditure The bi modal distribution reflects the maximum claim for families with different numbers of children MARGINAL TAX RATE FACILITY Introduction and Overview Guide Page 39 SPSD M Version 9 2 The marginal tax rate is defi
58. first running a 1992 base case to observe the level of federal tax revenue obtained when manufacturer s sales taxes are collected and then creating a variant case which sets manufacturer s sales taxes to zero The facility operates by iteratively executing a black box version of the SPSM and manipulating the parameter files between runs based on the values found in the default tables REFERENCE VARIABLE FACILITY Values of any number of modeled variables produced in an SPSM run can be accessed in a subsequent run by using SPSM reference variable facility The facility permits the simultaneous comparison of any number of simulation scenarios The most common use of the Reference Value Facility is to access user variables produced in a previous run The reference value facility can also be used to replace the values of SPSD variables The user may use this feature to for example model a behavioural response of persons responding to a shortfall of Ul benefits by increasing their employment income Example Simulations This section contains several worked examples of SPSD M use They should be performed by the user at his or her microcomputer in the sequence presented here The examples provide guidance on using some of the many SPSD M features The complete dialogue for producing each example run is given below Inputs that the user typed have been shown in bold Lines have been numbered in square brackets and notes on these inputs have been int
59. g SPSM 1 C3 MKDIR SPSMTEST 2 C gt CD SPSMTEST 3 C SPSMTEST gt SPSM 1 Any analysis with the SPSM will usually begin by setting up a working directory for the parameter files and output files pertaining to the particular analysis in this case named SPSMTEST 2 The working directory should then be made the current directory 3 The executable program is invoked by typing SPSM Remember that the MS DOS search path must contain spsm win32 At this point the SPSM program is loaded into memory and execution begins The user will first see a copyright message particular to their version of the SPSD M briefly displayed on the screen Shortly thereafter the message disappears and the SPSD M start up screen appears Welcome to the Statistics Canada Social Policy Simulation Database Model SPSD M 4 Program Version 9 0 5 Database Adjustment Algorithm Standard 6 Standard Tax Transfer Algorithm Version 9 0 88 05 7 Alternate Tax Transfer Algorithm none aucun Compiler Operating System MSC 12 0 WIN NT Copyright c Minister of Industry 1984 2001 Press F1 for context specific help Press Ctrl Break to terminate LLILLlll2lllllll2l l 2 2 Control Parameters 8 Enter name of input control parameter file gt spsd ba92t Loading control parameters from spsd ba92t cpr 9 Enter specification for generating output files gt test1 Generating default output file
60. g output from SPSM standard and custom tables In addition statistical software packages such as PC SAS and SPSS are useful for subsequent analysis of simulation results USING THIS GUIDE Certain portions of this guide should be read by everyone who will be using the SPSM for research purposes from junior programmers to senior analysts e f you are unfamiliar with the SPSD M and never plan to use the program or analyze and interpret the results directly you can confine your reading to this introductory chapter e Persons familiar with microsimulation and MS DOS can skim certain paragraphs and sub sections but should follow the guide as structured The SPSD M is a complex static microsimulation model and must be understood thoroughly if one is to make correct inferences The computer program itself however is simple to use and results can be produced very quickly with it Overview of the SPSD M presents the fundamental concepts which must be understood prior to conducting any research with the model This includes a description of all the basic parts of the model and their relationships to one another By reading this the user will also become familiar with the specific terminology which is used throughout the documentation The section titled Software and Facilities describes the use and function of the SPSM user facilities These facilities allow users to modify parameters interactively screen or select records produce default and cust
61. gt 1202 gt go Writing out changed tax transfer parameters to factc1 mpr Changing INPVARMPR from SPSD ba92 mpr to factc1 mpr SPSM started on Mon Sep 17 09 28 51 2001 SPSM completed on Mon Sep 17 09 28 53 2001 Elapsed time 00 00 02 Number of households processed 3990 Number of persons selected 11954 Variant consumable income 000 000 365958 4 Increase from base 000 000 1724 4 Control parameter file factc1 cpr Table output file factc1 tbl Variant tax transfer parameter file factct mpr Variant results file factc1 mrs A descriptive label of the variant results is given as usual At this point all of the parameters identified earlier must have their values set to 0 The federal family allowances are turned off first followed by the specific Alberta and Quebec federal family allowance configurations Quebec family allowances vary by the number of children The settings for Introduction and Overview Guide Page 62 SPSD M Version 9 2 this table parameter are contained in the file SPSM EXAMPLE QFFSL MPI It is read in here to save the user the trouble of editing all the entries of this table lookup parameter 20 The last parameter to change is the doubling of child tax credit benefits The model execution is then initiated with the go command The following is a listing of the contents of the table result file for the FACTC1 run factc1 tbl SPSD M Version 9 00 Bas
62. h Household sequence number 62 idefseq Economic family sub sequence number 0 O 0 idcfseq Census family sub sequence number 0 O 0 hdprov Province ssse 6 idage Age esses 53 50 18 idsex SeX xS usos 0 1 0 idcfrh Relationship to census family head 0 1 2 immmkt Market income 36457 3750 4274 immtran All transfer income 01222 0 An ASCSTYLE value of 2 produces a report designed to be read using a spreadsheet import function while 3 produces a report designed to be read using a spreadsheet or a database system An ASCSTYLE value of 5 produces a report designed to be converted into a compressed format that can be read by the SPSM This feature allows users to modify SPSD database files Below is an example of ASCSTYLE 5 for one household not all variables values are shown 000001 47 2172121341 00000000 4203111399123052100000000010 00 SAS Output Facility SSD SAS Institute produces a microcomputer version of their mainframe Statistical Analysis System with almost identical functionality SAS has extensive facilities for Introduction and Overview Guide Page 41 SPSD M Version 9 2 analysis reporting and manipulation of files of microdata In order to take easy advantage of these facilities SPSM incorporates an interface to the SAS PC system This interface produces self documenting binary files in the same format that SAS it
63. h single parent census families on average received 663 in combined federal provincial family allowances they lost only 572 in consumable income as a result of eliminating benefits The 91 difference is accounted for by the reduction in taxable income and consequently taxes paid Therefore single parent families receive on average an after tax benefit from federal and provincial family allowances of 572 The 91 change in consumable income is actually accounted for by a mix of income and commodity taxes The last two columns split the effective tax into its two components Note that a portion of the effective tax will be affected by changes in other entitlements as a result of eliminating family allowances For example child tax credit benefits would be slightly higher due to the lower incomes without family Introduction and Overview Guide Page 55 SPSD M Version 9 2 allowances The second row of Table 1U gives results for two adult nuclear families Two adult families have higher average incomes and so have higher marginal tax rates Consequently while their average family allowances are significantly higher than single parents their average net benefit is slightly lower Table 2U Selected Quantities for Nuclear Families by Base disposable income group po dece du qe T eE Base Mean Avg Census SCF disposable Change children Families Records income Disposable per family group Income TM
64. he following section for a description of the four kinds of output files A CMD file contains a sequence of responses to SPSM prompts corresponding to an actual user dialogue This file can then be specified as an argument when invoking SPSM SPSM will then take input from the CMD file rather than the keyboard The relationships of all these files to the structure of the SPSM is shown in the following diagram File types with their extension are shown in ellipses Introduction and Overview Guide Page 30 SPSD M Version 9 2 Inout DIALOGUE Pararieters DIALOGUE NA SU UA C Adjust Data f S j T T Algorithm oa x P4 S Adjust Database CO MTAX Calculate Base Results Calculate Variant Results Record Selection User Variables Output Facility DADADS PARMS TABLES ASCII RESULTS Vemm Output SPSD M Files and Relationships This diagram can be used as a handy reference while becoming familiar with the files and naming conventions Note that two tax transfer parameter files are input for base and variant results calculations Also the SPSM results file can only be used to give base results and can only be written using the variant results Implicit in the diagram is the output of modified database adjustment and tax transfer parameter files during the dialogue SPSM User Facilities The SPSM user facilities create a powerful and flexible environment fo
65. ily Allowances and were ineligible for a Child Tax Credit both before and after the change The low 596 reduction rate together with the increased amount per child combine to make middle income families eligible for the credit A number of features of the SPSM model have been demonstrated in this example The results once examined and explained tend to be easy to understand intuitively once one has grasped which tax transfer programs are interacting and how each of these interacting programs work While the qualitative conclusions could be derived independently perhaps with considerably more thought than the ten minutes required to run the model the SPSM model adds a quantitative dimension to the analysis For playing around with ideas the 5 model appears to be quite adequate and saves time SESSION 4 NEW EARNED INCOME TAX CREDIT SIMULATION This final tutorial session shows a more advanced use of the SPSD M in black box mode It is not necessary to proceed with this example before conducting research with the SPSD M Those readers who are comfortable with the use of the model thus far are invited to try this last session In this session we will be testing the fiscal impacts on households and governments of introducing an earned income tax credit EITC program The testing of a new program would normally be a glass box application requiring re compilation of the Introduction and Overview Guide Page 72 SPSD M Version 9 2 model Ho
66. imple one word command The new system may then be used in black box mode The user need not be a C language expert in order to use glass box mode but should be experienced with Introduction and Overview Guide Page 4 SPSD M Version 9 2 some high level computer language The 1997 SPSD M may be projected to represent a variety of different years using a method of static aging A set of database adjustment algorithms and alternate weight files are included with the model to this end The algorithms allow the user to grow or project the value of the money items on the base year database to some future year The user can specify a series of growth rates which are then applied by the model In most cases a single scaling factor is applied However for certain important or complex variables more complicated rates are applied For example employment earnings are grown by rates specific to province and sex of the individual These algorithms can be used independently of or in conjunction with annual demographic weight files included as part of the database Users may add their own adjustment routines to augment or even replace the supplied adjustment algorithms For example if users wish to grow employment income at rates differing by age and sex they could set the existing employment income scaling parameters to one and create a new income scaling algorithm SOFTWARE AND FACILITIES The SPSD M comes with a powerful software support environme
67. isp eitc label dispinc Disposable Income 11 gainer eitc gt 0 label gainer Received EITC Flag Gainer 12 Nochange eitc 0 label nochange Unaffected by EITC Flag 13 agegrp split idage 20 64 label agegrp Age 14 Empigrp split idiemp 0 8000 12000 24000 15 Dispgrp split immdisp 5000 10000 15000 20000 25000 30000 35000 40000 45000 label dispgrp Base disposable income group 10 Because the definition of immdisp disposable income is not modifiable in black box mode we will need to create and label a new variable dispinc which shows disposable income after the new tax credit 11 gaineris true 1 if the individual gained from the EITC Otherwise the value is zero 12 nochange is true 1 if the individual was unaffected by the EITC Otherwise the value is 0 13 We want to examine the results across three different classification dimensions These new variables allow us to validate the SPSM implementation of the new EITC program and to measure the distributive effects of that program The agegrp variable breaks the sample into three groups according to age These groups are use to produce custom Table 1U 14 The empigrp variable has five different values depending on the level of employment income This allows us to check that nobody is receiving benefits if family income is above the breakeven value for the EITC 15 The dispgrp variable has ten different values depending on the level of disp
68. its are reduced by 10 for each additional dollar of earned income The above specifications can be represented by the following graph Introduction and Overview Guide Page 73 SPSD M Version 9 2 Benefits 1 500 1 200 1 000 N 15 2 500 8 000 12 000 Family Income 24 000 Figure 9 Earned Income Tax Credit Benefit Structure The four figures presented on the following pages contain a listing of a file named eitc cpi which will be used to perform the analysis Numbered comments are used to explain the file s contents 1 HHHH eitc cpi Include file for Tutorial Session 4 THE S d THE Purpose This file contains the necessary control parameters to simulate HH a new earned income tax credit The description of this new THE program and explanation of parameters is contained in the HH Introduction and Overview Guide under the heading of Tutorial TH Session 4 THE THHBBHHHHHHHBHHHBHHBHEHHBHHBHHHHHBHHBHHBHHBHHHHBHBNHHHBHHBHBHHHBHBHHHBHHHEHHER HHHHBHHHHHBHHHHHBHBHHSHE HHH 2 2 1 7 Record selection facility HHH SELFLAG 1 Selection facility activation flag SELUNIT 2 Selection facility family level SELSPEC Selection specification idage lt 21 amp amp idcfrh 2 1 The comment block at the beginning of the file provides information on the purpose of the file contents as well as the name of the file These comment blocks are useful in parameter include files but should not
69. l and all analysis variables will be rolled up to that level However the classification variables must be calculated for the desired family level of table Thus for a NF table the user would group NF immdisp rather than immdisp Also the age would refer to the head of the nuclear family The distributional analysis facility is used to explore the shape of the distribution of new EITC benefits The level of analysis is set to individual benefits DISTUNIT 0 Because DISTZERO is set to 0 individuals receiving no EITC benefits are excluded from the distributional report The requested breakpoints DISTP are altered due to the truncated nature of the distribution Introduction and Overview Guide Page 77 SPSD M Version 9 2 This file can be found in the SPSM EXAMPLE directory The whole SPSM run can be invoked in batch mode by typing the following command at the MS DOS prompt SPSM spsd ba92t eitc y read spsm example eitc cpi go n n n You will see the entire dialogue appear on the screen exactly as if you had typed the commands in response to the prompts Let us now turn to briefly examine the outputs of the run which have been written to the eitc tbl file Table 1U Selected Quantities for Individuals by Age Age ToMmMMMMMMMeMMMMUS TM Quantity Min 20 21 64 65 Max All TM To T Earned income tax credit M 0 0 1201 1 3 7 1204 7 Received EITC Flag Gainer 000
70. l be produced by reading an existing SPSM results file Introduction and Overview Guide Page 49 SPSD M Version 9 2 6 The name of the results file created in the previous example Session 1 is specified Recall that this contains values for the 1992 tax transfer system for four simulated variables disposable income consumable income federal family allowances and provincial family allowances THE 2 1 4 Variant Information 7 VARDESC 1992 Tax System No Family Allowances UVARFLAG 1 8 UVAR 9 deltacon immicons immicons label deltacon Change in Consumable Income deltadsp immdisp immdisp label deltadsp Change in Disposable Income 10 totalfa _imffa impfa label totalfa Total Family Allowances 11 iskid idage lt 18 label iskid Child 12 dispgrp split NF immdisp 5000 10000 15000 20000 25000 300 00 35000 40000 45000 label dispgrp Base disposable income group 7 The variant run which will be compared to the status quo 1992 system is labeled 8 We will use the user variable facility to examine model results of interest 9 We will want to report on the difference in disposable and in consumable income between the base and the variant The difference in consumable income will be stored in the user defined variable deltacon The expression should be read as deltacon equals variant consumable income minus base consumable income Note that the leading underscore refers to the base variable value
71. m FATBPI Family allowance take back phase in FATD Family allowance turndown income FARR Family allowance repayment rate And two pages later the child tax credit parameters are listed as follows Federal Child Tax Credit CTCOPT Federal child tax benefits calculation option CTCPC Child tax credit per child CTCTD Family income child tax credit turn down CTCRR Child tax credit reduction rate CTCERF CTC child care expense reduction fraction CTCREF CTC post reform rules flag CTCSUP CTC young child supplement CTCTUR CTC take up rate table CTCINC CTC family income scaling factor The parameters to be changed have been identified but their new values still need to be determined The elimination of the federal Family Allowance program should pose no problem since this can be effected by setting all of the relevant parameters to zero The parameter QFPSL since it represents an extra Provincial contribution above and beyond the Federal Family Allowance will be left unchanged The FAFLAG parameter will not be changed because if it were set to O then provincial family allowances in Quebec would not be calculated Evidently the parameter Introduction and Overview Guide Page 57 SPSD M Version 9 2 CTCPC must be increased but the amount of increase required for consumable income neutrality is not clear An iterative approach involving more than one SPSM program run will be required The strategy for accomplishing this is as follows Step
72. mable income on the other hand is defined as disposable income less commodity taxes embodied in household consumption As such it is not a post all tax concept because not all taxes are accounted for when determining consumable income For example municipal property taxes and provincial land transfer taxes are not included in the calculation The SPSD M built in tables and standard reporting facilities define gainers and losers in terms of consumable income If users wish to express their analyses in terms of disposable income two approaches are possible First the user can disable the calculation of commodity taxes and cause consumable income to be set equal to disposable income Alternatively the user can specify custom tables using the X tab facility which are defined in terms of disposable income SPSM FILES Files are maintained using a MS DOS directory structure which follows certain conventions The SPSD M uses a great number of different types of files All Introduction and Overview Guide Page 29 SPSD M Version 9 2 SPSM executable programs are stored in the spsm win32 directory For example the full pathname of the executable SPSM program is spsm win32 spsm exe Database files and supplied parameter files are found in the spsd directory The use of directory structures and strict naming conventions have been used in order to avoid confusion The following table gives the file extension naming conventions used for SPSM input
73. n The run can thus be duplicated using the output control parameter file in a subsequent run Certain of the control parameters are informational and are not directly modifiable by the user An example is ALGDESC that displays which version of the tax transfer algorithm was used for the standard and alternate tax transfer algorithms If the user has modified database adjustment parameters or tax transfer parameters in the user dialogue an output parameter file of the appropriate type including the specified changes is produced Case by case results are useful for a number of reasons They can be used as input to other software programs If for example a user wished to calculate gini coefficients or perform a regression analysis on SPSD M variables he or she would create a dataset containing only the variables of interest for each household or any Introduction and Overview Guide Page 28 SPSD M Version 9 2 other level of analysis That file would be read and processed by other analytical software with regression and sorting capabilities such as SAS SPSS GAUSS etc Case by case results can also be used as input to the SPSM to provide base result variable values Finally case by case output can be printed or viewed on the screen for debugging or other analytical purposes For example if a user wished to examine in great detail the five households who had lost the most disposable income in a comparative simulation he or she would first isola
74. n 2 SESSION 2 CUSTOM TABLES SCREENING RECORDS USER ANALYSIS VARIABLES In this example the results file created in Session 1 is compared with a variant in which all family allowances are eliminated The time required to run the SPSM is roughly cut in half by reading a base results file rather than calculating base results As a further time saving measure the simulation run will only be conducted for census families having children because those are the only recipients whose economic position will change as a result of the elimination of family allowances The Introduction and Overview Guide Page 48 SPSD M Version 9 2 run will produce two custom tables which make use of several features The example makes use of an auxiliary parameter include file session2 cpi which was created using a text editor and is supplied in the directory SPSM EXAMPLE This file will be read in during the interactive dialogue This technique is one way of organizing your work and minimizing repetitive dialogue sessions This example will first examine that input file the model dialogue and finally the results tables The following three figures provide a listing of the file session2 cpi It was created by editing the base 1992 control parameter file Only those groups of parameters which will differ from the base are retained along with their labels The mark indicates comments HHHH 1 session2 cpi Include file for tutorial Session 2 TH d
75. n SPSM program run Once the user has started the model and provided a complete set of parameters the simulator phase begins The simulator phase operates in five sequential steps Each of these steps is described briefly below Introduction and Overview Guide Page 21 SPSD M Version 9 2 Adjust Data Calculate Base k Scenario Simulator Calculate Variant Scenario Calculate User Variables Selection and sa m Output Routines Database Adjustment The SPSD is based on surveys and administrative data collected for the 1997 calendar year Adjustment of this data may be required for several reasons The user may wish the incomes to represent the incomes in a more recent year such as 1999 In this case the user would then supply estimates of growth rates for income by source between 1988 and 1999 The user may also wish to make adjustments for simulation purposes in the 1997 year For example known under reporting of alcohol expenditures may be treated by scaling each household s alcohol expenses or a deduction item could be scaled down to zero to represent a restriction of a certain tax measure Finally the user may wish to adjust the weights of individuals and households to reflect alternative demographic assumptions Prior to the calculation of any taxes or transfers each dollar denominated database variable is multiplied by a single corresponding database adjustment parameter There are two key exceptions to this Employment i
76. n method CTDFLAG 0 Commodity tax detailed calculation flag TARGETYEAR 1992 Year of analysis 44 Do you wish to modify any variant tax transfer parameters gt y gt mprdesc Description of tax transfer parameter file 45 MPRDESC gt FACTOS variant no federal FA 906 25 CTCPC gt ctcpc Child tax credit per child 46 CTCPC gt 906 25 Introduction and Overview Guide Page 69 SPSD M Version 9 2 47 43 44 45 46 47 gt go Writing out changed tax transfer parameters to factc3 mpr Changing INPVARMPR from factc1 mpr to factc3 mpr SPSM started on Mon Sep 17 09 47 36 2001 SPSM completed on Mon Sep 17 09 47 37 2001 Elapsed time 00 00 01 Number of households processed 3990 Number of persons selected 11954 Variant consumable income 000 000 364301 5 Increase from base 000 000 67 4 Control parameter file factc3 cpr Table output file factc3 tbl Variant tax transfer parameter file factc3 mpr Database adjustment as mentioned before is not being used for this tutorial session We will need to modify the tax transfer parameter file which was specified as FACTC1 MPR Correct file documentation is maintained as usual The only other tax transfer parameter which will change is CTCPC the value of the child tax credit per child We will give it the value of 906 25 as determined earlier Having made all the specifications for this final run we ca
77. n one of the parameter files by answering yes to the appropriate question the gt prompt indicating parameter editing mode is displayed Four commands can be used in response to this prompt in any of the three stages Introduction and Overview Guide Page 32 SPSD M Version 9 2 GO Terminate modification of this set of parameters and continue with the next stage in the user dialogue LIST List all valid parameter names in the current parameter set on the user s display terminal READ Open and read a user specified file containing values for one or more parameters name Engage in a dialogue to allow the display and modification of the parameter named name Changing Parameters The SPSM uses several types of parameters Different types are checked for validity using different rules The following is a description of the parameter types currently defined FLAG A flag controls whether or not an algorithm or program or calculation is performed OPTION An option parameter allows the choice of one of two or more options SCALAR Scalar parameters take a single numeric value with or without a decimal point STRING A string parameter is a single line or multiple lines of text VECTOR A vector is a single column of numbers The first value is the number of values which follow LOOKUP TABLE A lookup table contains a single value followed by a set of three columns The first column represents the input value such as taxabl
78. n proceed to the execution of the simulation The following is a listing of the first user specified X tab table found in the FACTC3 TBL file Introduction and Overview Guide Page 70 SPSD M Version 9 2 SPSD M Version 9 00 Base Description Current Values for 1992 Variant Description FACTC3 variant no federal FA 906 25 CTCPC Sample 0 0510 AGENAME Standard Table 1U Selected Quantities for Census Families by Province TM 4 Province Federal Net Provincial Change in Balance Net Balance Consumable Increase Increase Income M M M Newfoundland 10 3 8 4 1 9 P E I 1 0 1 3 0 3 Nova Scotia 7 5 8 0 0 5 New Brunswick 8 0 6 9 1 0 Quebec 97 8 136 6 38 8 Ontario 85 5 98 8 13 4 Manitoba 285 7 1 11 4 Saskatchewan 12 4 18 5 1 1 Alberta 9 5 20 6 30 2 B C 30 4 27 9 2 4 4 7 All 271 8 339 2 67 4 4o 4 7 4 4o A number of features of our example scenario can be seen in Table 1U The first obvious one is the effective transfer of about 272 million from Provincial Governments to the Federal Government despite the overall neutrality of the scenario in terms of consumable income This is a direct result of the elimination of the taxable Family Allowances This elimination results in a decrease in provincial
79. n the FACTC1 run We shall make a subsequent Introduction and Overview Guide Page 68 SPSD M Version 9 2 modification to the value of CTCPC in this file in the tax transfer parameter editing phase of the dialogue 41 The settings from the include file for the user defined variables and the X tab specification are read in at this point The XTFLAG parameter is set to 1 in this file to ensure that a custom table will be produced 42 Changes to the control parameters for the FACTC1 run are now complete and we can proceed to the next phase of the dialogue Loading database adjustment parameters from SPSD ba97_92 apr Current values of selected database adjustment parameters GFISENF 10 Growth Factor Self employed income non 0 74940 0 89140 0 85640 0 79350 0 87880 0 77390 0 79200 0 75870 0 79820 0 85950 GFIEMP 10 Growth Factor Employment income 0 96450 0 96450 0 99370 0 99370 0 99370 0 99370 0 96230 0 96230 0 95560 0 95560 0 94580 0 94580 0 97530 0 97530 0 88710 0 88710 0 90040 0 90040 0 95650 0 95650 43 Do you wish to modify any database adjustment parameters gt n EE Tax Transfer Parameters Loading variant tax transfer parameters from spsd ba92 mpr Loading variant tax transfer parameters from factc1 mpr Current values of selected variant tax transfer parameters MPRDESC X ba92t No Federal FA CTC 2 CTFLAG 1 Commodity tax activation flag CTOPT 2 Commodity tax calculatio
80. n the tax transfer algorithms A complete list of parameters organized by subject area can be referenced in the Parameter Guide SPSM parameters provide instructions to the simulator in the following way Introduction and Overview Guide Page 26 SPSD M Version 9 2 Tax Transfer Parameters Income Payroll Commodity Taxes Scenario Database Control Adjustment Taxes amp Cash Transfers Get Data Calculate Base Scenario Calculate Variant Scenario Calculate User Variables Selection and Output Routines The Relationship of Parameters to the Simulator Note that control parameters affect every step of the simulation whereas database adjustment parameters are used only once Two sets of tax transfer algorithm parameters are specified one for calculating base results and one for calculating variant results Each of the three types of parameter files are now discussed in turn Control Parameters A control parameter file contains specifications for parameters used to control SPSM general operations It gives information for such things as input and output file specifications record selection requests cross tabulation and other user facilities It does not give information on the parameters used to specify database growth or parameters of the tax transfer system This information is provided by separate parameter files The majority of control parameters are used to control the SPSM user facilities The control parame
81. nal retail stage Rates calculated by the COMTAX model can be input as parameters into the SPSM to obtain estimates of the commodity taxes paid both directly and indirectly by any given household The data retrieval and reporting software are configured as facilities accessed through the model They allow the user to produce formatted output and to perform specific kinds of analysis The user documentation is extensive and comprehensive It is divided into three manuals with a number of guides in each CAPABILITIES AND FEATURES The SPSD M algorithms operate on a database created using survey and administrative data from 1997 The SPSD M comes with the necessary algorithms and parameters to simulate the tax transfer systems in that year The SPSD M also performs projections to all other years from 1988 through 2005 The following is a list of the major programs which can be simulated with the SPSM Users can add their own algorithms to this set Introduction and Overview Guide Page 3 SPSD M Version 9 2 Personal Income Taxes e 1988 to 2005 taxation year algorithms e Major deduction items e Payroll deductions CPP QPP UI contributions e Provincial tax algorithms Cash Transfers e Unemployment Insurance Benefits Employment Insurance Detailed treatment of eligibility and benefits by type regular maternity fishing sickness retirement and phase regular labour force extended regional extended Old Age Security benefits i
82. ncluding partial benefits Guaranteed income supplement benefits including partial benefits Six provincial GIS supplement programs Family Allowance benefits federal and provincial Child Tax Credit Federal Child Benefit Sales Tax Credit ommodity Tax Model Twelve different tax types accounted for including Federal Import Duties Federal Manufacturer s Sales Tax Provincial profits on liquor commissions Federal Fuel Taxes Provincial Fuel Taxes Provincial Retail Sales Taxes e Cascading effects between taxes e Determination of household expenditures net of implicit commodity taxes eoe o gt gt o o o The SPSM is written in the C programming language Changes to the tax transfer model algorithms can be made in two distinct ways the black box or the glass box mode The majority of SPSM analyses operates in black box mode When using the model in the black box mode a user would specify changes to one or more of the over 800 tax transfer parameters during a user friendly dialogue In this mode the user can change levels and rates affecting various benefits taxes and eligibility requirements and can also specify a sub population that will be modeled The glass box mode is designed for use by persons who wish to develop entirely new algorithms or to change existing algorithms Glass box users can write a new or modify an existing algorithm in the C programming language and then recompile the system using a s
83. ncome comprising about 70 of total income is given more rigorous treatment due to its relative importance in the overall tax transfer system Employment income is therefore adjusted by industry specific growth rates CPP QPP income is adjusted by differential rates according to age groups This allows a simulation of the phasing in of that program Users should refer to the Parameter Guide for further details on specific adjustment parameters Calculate Base Results and Variant Results Microsimulation results are usually analyzed by comparing the results of one specified scenario with those of another For example if one was interested in examining the impact of extending Unemployment Insurance maternity benefits from Introduction and Overview Guide Page 22 SPSD M Version 9 2 15 to 20 weeks two scenarios would have to be calculated The first scenario would be the tax transfer system as currently legislated let us call it Scenario 1 The second scenario has the exact same tax transfer system with the exception that Unemployment Insurance maternity benefits are extended to 20 weeks Let us call this second scenario Scenario 2 This could be done by performing two separate program runs and then comparing the outputs from the runs to determine the impact However the SPSM is capable of performing the comparison in a single program execution This is accomplished by applying the simulation algorithms to a given household twice in the same pr
84. nd utilities designed for use with SPSD M such as the spreadsheet interface and parameter file comparison utilities The currently provided auxiliary tools take the form of other executable programs or Lotus 1 2 3 and Symphony spreadsheets SPSD M REFERENCE MANUAL There are three main ingredients to the SPSD M which are manipulated in various ways to create various tax transfer scenarios the database the algorithms and parameters This manual provides detailed information on the components of each of these three ingredients Algorithm Guide The Algorithm Guide is designed primarily to provide SPSD M users with an understanding of the specific personal tax and cash transfer programs which are simulated together with specific information on how each program was implemented An attempt is also made where possible and practical to guide and caution the user Introduction and Overview Guide Page 81 SPSD M Version 9 2 where necessary as to the interpretation of the results The guide is intended for both black box and glass box users Parameter Guide This document provides an explanation of all parameters including control parameters that are part of the SPSM This includes a detailed description of how the parameters are used their values for the various parameter files provided and wherever possible the published source where the values were obtained Variable Guide This document contains reference information on each SPSD M va
85. ned as the proportion of an extra dollar of income that is taxed It is a useful concept because it measures the extent to which incentives to obtain additional income are reduced by the tax transfer system SPSM provides a facility to compute marginal tax rates When the facility is activated the tax transfer system is applied twice to each household once to the original incomes and once to the incremented incomes and the resulting change in consumable income is noted The definition of marginal tax rate given above is not complete To calculate a marginal tax rate one must in addition specify the source of income being incremented the amount of the increment to income and which individuals are to receive the increment The change in consumable income at the individual level can then be aggregated to produce marginal tax rates at different family levels of analysis Marginal tax rates differ depending on the family level of analysis Consider a married couple where one person has zero income If 500 is given to the individual with no income that person s marginal tax rate would be zero The consumable income of the person s spouse would decrease however due to a reduction in the married tax credit deduction Hence the marginal tax rate of the spouse would compute to an infinite value since the spouse s taxes have increased even though he she received no additional income Considered as a family however the marginal tax rate would evaluate
86. nly limited control over the format and contents of these tables The user specifies a level of analysis which applies to all of the built in tables He or she can also specify the variable as well as the cut points to be used in determining the columns of Table 2 Thus the user can easily create a table of results by age A separate utility allows any of these built in tables to be transformed into a format which is readable by Excel spreadsheets An Excel spreadsheet which produces auxiliary statistics from these tables is also included Purchaser s of the Full SPSD M package may refer to the Tools User s Guide for further details User specified X tab Facility The designers of the SPSM recognize that the built in tables will not always meet the analyst s needs Thus a powerful facility that allows a user to create specific customized tables has been provided A separate auxiliary guide XTab User s Guide presents detailed documentation on the X tab facility in a comprehensive format that is also suited to those learning to use the facility An individual tabulation request consists of multiple components It typically begins with the specification of the relevant unit of analysis i e individual IN nuclear family NF census family CF economic family EF or household HH If the tabulation request does not include a unit specification a default unit of individual IN applies The remainder of the tabulation request consi
87. no benefits are paid to individuals with over 24 000 of earned income The 297 3 million in EITC benefits paid to individuals having no employment income appears to be an error Recall however that the program as specified is based on family employment income This means that if an employed person earning between 1 and 24 000 had a non working spouse that spouse would be eligible for individual benefits It is apparent that this version of the EITC is not a pure individual work incentive program Notice that the average benefits for the 8001 12000 employment income group is 1 041 Some readers may have expected the number to be 1 200 or maximum benefits This would be the case if the table were restricted to unattached individuals Again this is not the case as the table is based on the income of Introduction and Overview Guide Page 78 SPSD M Version 9 2 individuals but the benefits are based on family employment income The 1 041 is short of the maximum due to married couples who have more than one earner and whose combined income is less than 24 000 In these cases the income of the second earner moves the family employment income into the range where benefits are reduced Table 3U Selected Quantities for Individuals by Base Disposable Income Group o To o 4 To Base Earned Received Unaffected SCF disposable income tax EITC Flag by EITC Records income credit M Gainer Flag 000 group 000
88. nt designed to enhance the range speed and ease of policy analysis in both black box and glass box modes The comprehensive and flexible set of reporting options allow for detailed examination and verification of the database algorithms and results The software has been designed to work in an integrated fashion at different levels of analysis e g household census family individual A user friendly dialogue guides the analyst through all the required and optional adjustments of the parameter files An on line help facility provides information on the use of the model dialogue as well as a handy reference to parameter lists Some highlights of the model are e The user expression facility allows the user to create name and access in subsequent runs new variables that can be used for reporting purposes without the need for any C language programming e The generalized record selection facility allows the user to specify a database subset in terms of any combination of database or modeled variables For example a user could create a condition with an alternate definition of income and then select all economic families with less than a specified amount of that income e Marginal tax rate analyses can be performed in a single run of the model e Two distinct tax transfer systems can be run simultaneously The two systems can be compared using any combination of variables in the same run Base and variant runs may be differentiated on th
89. of Ul on the SPSD are given Included are data related to the start date of a claim the type of claim and weeks on UI in the various phases e g Regional Extended Benefits Variables come from matching EFC UI claimants with a 10 random sample about 400 000 records in the base year of Human Resources Development Canada administrative data files on individuals in receipt of UI El Tax related Variables To complete a tax form one requires certain individual non income variables These include RRSP contributions tuition fees medical expenses and charitable donations These variables come from a stratified sample of approximately 350 000 Revenue Canada T1 individual income tax returns the same sample that underlies Revenue Canada s annual Green Book publication Consumption Pattern Variables Each household on the SPSD has an associated expenditure pattern giving household expenditures by 48 distinct types of commodity The Family Expenditure Survey FAMEX collects this detailed information on expenditure patterns on about 10 000 households Household Weights Household weights are stored on separate files There is a series of nine sets of weight files corresponding to population estimates for the years from 1984 to 1997 Introduction and Overview Guide Page 19 SPSD M Version 9 2 The database variables mentioned above have been arranged into fourteen distinct groups A full list of the available variables ordered by group is given
90. ogram run but with different parameter files One such application is called the base and produces base results while the second application is called the variant and produces variant results The first run Scenario 1 would be calculated using the base set of parameters while Scenario 2 would be calculated using the variant parameters The results of any single simulation scenario are derived by a series of functions which sequentially calculate certain taxes and cash transfers as required by the tax transfer system For example family allowances must be known before calculating net income The sequence of these functions is controlled by a program called the driver The following figure represents the standard algorithm driver Changes to the standard or alternate algorithm driver can be made in glass box mode The following figure shows the structure of the standard tax transfer algorithm driver in more detail Adjust Database Calculate Base Scenario Calculate Variant Scenario Calculate User Ul OAS FA Federal Taxes Provincial Taxes GIS SPA GIS T Refundable Tax Credits Commodity Taxes Variables Reporting Variables Selection and Output Routines iud n C Introduction and Overview Guide Page 23 SPSD M Version 9 2 Tax Transfer Algorithm Driver Structure The first three functions simulate the Unemployment Insurance Old Age Security and Family Allowance programs whose benefits are determined b
91. ogramming The user defined variable facility UVAR allows the user to create new reporting variables in black box mode It allows the user to perform many analyses that would otherwise require programming changes to SPSM A new variable can be used just like any other SPSD M variable SPSD M variables have built in labels which are used by the SPSM reporting and output facilities The user can supply a label to an analysis variable or it will use defaults SPSD M user variables also have an associated output precision which gives the number of digits after the decimal point to use when formatting the variable in reports There are many applications for these variables Counts of individuals or family units with specified characteristics take up rates averages percentages and even new benefit functions can easily be incorporated in black box mode CREATING TABLES Automatic Parameter Reporting Comparisons between parameter files are essential in order to isolate exactly what has changed between any two simulation scenarios The SPSD M provides two tools for this The first is a tax transfer parameter comparator whose output is stored in the TBL file and is activated with using a control parameter This report shows tax transfer parameter differences between base and variant scenarios A more sophisticated parameter difference report can be obtained by using the compparm utility which is documented in the Tools User s Guide This is an au
92. om tables and distribution reports create new variables and conduct marginal tax rate analyses After reading through the above sections the user should proceed to the four tutorial examples The examples are meant to be followed in order They provide general information on using the SPSM as well as details on specific SPSM features Many of the SPSD M user facilities are illustrated in these examples The final section entitled Guide to SPSD M Documentation gives an overview of the contents of the other two guides included with the full SPSD M package and provides some guidelines on their effective use OVERVIEW OF THE SPSD M The Social Policy Simulation Database and Model SPSD M is a tool designed to analyze the financial interactions of governments and individuals in Canada It allows the assessment of the cost implications or income redistributive effects of changes in the personal taxation and cash transfer system There are four basic elements to the SPSD M a database the SPSD a model the SPSM which is comprised of a set of simulation algorithms data retrieval and reporting software and user documentation Introduction and Overview Guide Page 2 SPSD M Version 9 2 The SPSD is a non confidential statistically representative database of individuals in their family contexts with enough information on each individual to compute taxes paid to and cash transfers received from government The SPSM is a static accounting model
93. omputer program such as a spreadsheet database system Introduction and Overview Guide Page 40 SPSD M Version 9 2 or other SPSM facility The variables whose values are to be displayed are given in a string parameter ASCVARS Each case that is output corresponds to the level of analysis given in another control parameter ASCUNIT The selection facility is generally used in conjunction with this facility since otherwise very large files could result The facility is especially useful for testing new scenarios For example if a number of households losing disposable income are noticed in a run that should only result in gainers the relevant information on an entire losing household can be printed out and examined A number of styles of output are supported An ASCSTYLE value of 1 produces a report designed to be easily human readable One household is output per page and one variable is output per line as shown below Text output facility style 1 lt page break gt hdseqhh Household sequence number 61 idefseq Economic family sub sequence number 0 O 0 idcfseq Census family sub sequence number 0 0 0 hdprov Province ssss 6 idage Age 28 24 0 Idsex Sex uisa O 1 1 idcfrh Relationship to census family head 0 1 2 immmkt Market income 27953 17350 0 immtran All transfer income 194 371 0 page break hdseqh
94. orical or classification variables An example of a class variable is idsex which records the sex of an individual Variables which take numeric values on which arithmetic operations can be meaningfully performed are termed analysis variables An example of an analysis variable is idiemp which records the employment income of an individual The distinction between class and analysis variables is an important one Certain of the SPSM facilities require that either a class variable or an analysis variable be specified in certain contexts Class variables can be converted into analysis variables or analysis variables to class variables using the SPSM s facilities to create user defined variables A further important distinction is database versus modeled variables Database variables as described above represent responses to survey or administrative reports and reside on disk in the SPSD file Modeled variables are those variables which are calculated during the actual simulation run and may also be subsequently stored in result files Because of the many different types of variables used by the SPSM extensive and strict naming conventions have been adopted For example all database variables stored at the individual level of analysis begin with the id prefix and all individual modeled variables begin with an im Further details on naming conventions can be found in the Variable Guide SIMULATOR The simulator is the second basic phase of a
95. osable income This allows us to examine the extent of progressivity of the EITC Introduction and Overview Guide Page 76 SPSD M Version 9 2 16 17 18 19 16 17 18 19 HHH 2 1 15 1 Built in tables HHH TOFLAG 0 Table 0 request flag HHH 2 1 15 2 User specified Tabulation facility HHH XTFLAG 1 X tab facility activation flag XTSPEC X tab specification IN eitc gainer S 3 nochange S 3 agegrp IN empigrp eitc eitc gainer L Average Benefits gainer S 3 nochange S 3 IN dispgrp eitc gainer S 3 nochange S 3 scfrecs THHE 2 1 15 3 Distributional Analysis Facility HHH DISTFLAG 1 Distribution facility activation flag DISTUNIT 0 Distribution facility family level DISTVAR eitc Distribution facility variable DISTSAMP 3000 Distribution facility sample size DISTZERO 0 Distribution facility zero inclusion flag DISTP 10 Breakpoints for histogram plot 1 5 10 20 30 40 50 60 70 99 The first built in table which is activated as a default in the ba92t cpr file is turned off here The X tab facility is used to evaluate the impacts of the new program The specification shown here produces three tables Note that the tables have been forced to be tabulated at the individual level of analysis IN In this type of application where benefits are being calculated on an individual basis the user may specify a table request at a higher family leve
96. oup files into categories by giving more information about what is in the file Directories Because fixed drives can store so many files you need a method of organizing files so you can efficiently store and retrieve data When working with MS DOS files a directory is used to organize groups of related files First you determine how your files are related and then you create a directory and add your files to it Directories are stored on the disk with the files Directory names must follow the same conventions as filenames MS DOS uses a hierarchical system called tree structuring that allows you to arrange files in directories First you group related files into a directory then group directories into a multilevel structure called a TREE This structure is easy to visualize if you think of the main directory as the root the subdirectories as branches and the individual files as leaves Subdirectories can contain data files or other subdirectories The tree grows as other directories are added For an example please refer to the figure showing the SPSM directory structure Note that the conventional notation starts with the tree root at the top with the branches growing downwards The root directory or simply the root is the main directory on a hard or floppy disk MS DOS creates the root directory each time it formats a disk The root directory cannot be deleted A subdirectory is a directory which is below and included in another
97. p of interest Seconaly it can allow SPSM to execute faster by restricting the number of households processed to those of interest Introduction and Overview Guide Page 35 SPSD M Version 9 2 A string parameter contains an expression which is evaluated for each individual in the household If the result of the evaluation is non zero the individual is considered to be selected If an individual is selected then everyone in the same family unit is also selected irrespective of the evaluated value of the selection expression for other individuals In other words if at least one person in the family unit is selected the entire family unit is selected To compute taxes and transfers correctly SPSM always simulates entire households Because of this selection has no effect on the values of any variables modeled or database at the individual level If however an examination of the database variables in the selection expression indicates that no individual in the household could possibly be selected then SPSM skips immediately to the next household This can result in a considerable decrease in SPSM execution time USER DEFINED VARIABLE FACILITY As already discussed the SPSM has two distinct modes of use termed black box and glass box The glass box mode provides the user with considerable flexibility to design new algorithms and create new variables but it can be somewhat complicated to use and requires some knowledge of pr
98. plore the broad effects of parameter changes SPSM incorporates a sub sampling facility Sub sampling can be controlled in two ways In the first method the user sets a control parameter to the sub sample desired SPSM will then read and process the SPSD until the desired sub sample has been attained In the second method the user presses the CTRL BRK key combination to interrupt the run after some fraction of SPSD has been processed In either case the sub sample actually obtained is recorded in the control parameter file and output tables are scaled by the reciprocal of this value Reading of the SPSD always proceeds sequentially but a relatively unbiased sub sample is nevertheless obtained because households in SPSD are by and large randomly ordered The ordering is not completely random however In order to improve the accuracy of results obtained when using sub sampling SPSD has been arranged as a running stratified sample Samples of set sizes the first 5 25 and 5096 of the file have been arranged to have representative provincial and household income distributions The demonstration version of SPSD is identical to the 596 stratified sample Selection Facility The SPSM selection facility provides a means to restrict the set of individuals and families which are processed by the SPSM output and reporting facilities Selection satisfies two distinct needs Firstly it allows the analyst to focus attention on a particular sub grou
99. r the analysis of various tax transfer scenarios All of the SPSM user facilities are controlled by parameters in the control parameter file This section describes the purpose of each facility and provides an overview of its use USER DEFINED VARIABLES One of the powerful features of the SPSM is the ability to evaluate user specified expressions These expressions allow users to create their own variables as logical Introduction and Overview Guide Page 31 SPSD M Version 9 2 and arithmetic manipulation of other SPSD M variables The expressions can be used in many of the SPSM user facilities described below Expressions can allow users to simulate entirely new tax and transfer programs totally within the black box mode The following elements may be combined to form valid user expressions Variables Any database modeled or user defined variables Operators Logical and amp amp or not comparative lt lt gt gt and arithmetic Constants Any number Parentheses To control the order of evaluation of expressions Level of analysis override The level of analysis of a variable For example the following expression could be used to produce a variable which represents the ratio of average employment income to total income for individuals aged 21 to 64 idage gt 21 amp amp idage lt 64 idiemp imitot The term to the right of the multiplication sign says that for any given individual take
100. read in a text editor The parameter specification should be typed exactly as shown The parameter Introduction and Overview Guide Page 61 SPSD M Version 9 2 14 15 16 17 18 19 20 17 18 19 specifies a table having three rows corresponding to base consumable income variant consumable income and the difference We are now ready to proceed to the next phase of the user dialogue No database adjustment parameters need to be changed for this example We will want to modify the variant tax transfer algorithm parameters so the appropriate response is y Notice that the base parameters are loaded first if requested but they cannot be edited in the user dialogue session Only variant parameters can be edited during the user dialogue gt mprdesc Description of tax transfer parameter file MPRDESC gt ba92t No Federal FA CTC 2 gt Stdfa Standard federal family allowance per child STDFA gt 0 gt afacl Alberta FA benefit per child aged 0 6 AFAC1 gt 0 gt afac2 Alberta FA benefit per child aged 7 11 AFAC2 gt 0 gt afac3 Alberta FA benefit per child aged 12 15 AFAC3 gt 0 gt afac4 Alberta FA benefit per child aged 16 17 AFAC4 gt 0 gt qfs Federal supplement per child 12 17 on Quebec family al QFS gt 0 gt read Name of file to read gt spsm example QFFSL MPI parameter QFFSL updated from file gt ctcpc Child tax credit per child CTCPC
101. riable The first part of this document is organized by subject area The second part consists of an encyclopedic reference to each SPSD M variable and is organized alphabetically The reference section is for determining the detailed definition characteristics and use of a particular variable Database Creation Guide This guide describes the general process step by step of constructing the micro statistically representative SPSD Introduction and Overview Guide Page 82 SPSD M Version 9 2
102. riables VARMETH 2 3t Method of creating variant variables INPVARMPR SPSD ba92 mpr Name of variant tax transfer Output Files OUTMRSFLAG 0 Variant results file creation flag OUTCPR factc3 cpr 3t Name of control parameter file out Do you wish to modify any control parameters gt y The FACTCS run will require new base variant settings a new set of reporting variables and a new X tab specification gt cprdesc X Description of SPSM run string CPRDESC gt 1992 Base vs No Federal FA CTCPC to 921 56 5 sample basmeth Method of creating base variables BASMETH gt 2 gt inpbasmpr Name of base tax transfer parameter file in s INPBASMPR gt spsd ba92 mpr gt inpvarmpr Name of variant tax transfer parameter file in INPVARMPR gt factc1 mpr gt read Name of file to read gt spsm example eg3xtab cpi parameter UVARFLAG updated from file parameter UVAR updated from file parameter XTFLAG updated from file parameter XTSPEC updated from file gt go No files will be overwritten with these control parameters Any further control parameter changes gt n Writing out control parameters to factc3 cpr As usual we will provide some documentation for the FACTC3 CPR file The base results should be calculated as the 1992 tax transfer system in the same was as the FACTC1 run The tax transfer parameters for variant results can be initially read from the FACTC1 MPR file created i
103. s in a subsequent run If base results are generated by reading in an existing results file the SPSM operates about twice as fast as if results had been calculated on the fly The SPSM results file is always written at the individual level of analysis and may contain any number of modeled variables TURNING POINT FACILITY The SPSM Turning Point Facility provides a means to analyze the points at which an individual household s marginal tax rate changes as its income increases The marginal tax rate is the proportion of tax that is paid on one additional dollar of income The turning points represent changes in the slope of the entire tax transfer function the marginal tax rate The facility computes the various income levels where an individual household given its characteristics would experience changes in its marginal tax rate and then outputs information on the current value of a number of variables at each of these levels of income The facility is useful for determining for example if a proposed means tested social assistance program will result in a disincentive to look for employment GOAL SEEKING FACILITY This tool is useful for answering what if type questions For example what would be the federal surtax rate necessary to achieve revenue neutrality given the federal Introduction and Overview Guide Page 42 SPSD M Version 9 2 manufacturer s sales tax is eliminated in 1997 It would be possible to answer this question by
104. scriptive notes have been included to the right of the table Introduction and Overview Guide Page 15 SPSD M Version 9 2 SPSD M Database 9 00 Base Description No base results Variant Description Current values for 1992 Driver Version 9 0 88 05 File SPSD ba92 mpr Sample 0 0510 AGENAME Standard Table 0 Results for Census Families Variable x1 000 000 TOTAL oom MM Family Units x1000 12597 3 12 6 Million Census families Persons x1000 28936 6 28 9 Million Canadians SCF Survey Records 1966 0 SPSD Records 4663 0 Income Base 364234 0 Income Variant 364234 0 364 Billion in Consumable Income Change 0 0 Total income less income and Number of Gainers x1000 0 0 commodity taxes Number of Losers x1000 0 0 No Change x1000 12597 3 Only one tax transfer scenario run Gainer s Gain 0 0 Loser s Loss 0 0 Total Income 529173 4 Market Income 450159 7 Wages and Salaries 348105 6 Self Employment Income 27107 6 Investment Income 38653 2 Other Income 36293 3 Transfer Income 79013 8 Total Tax 164939 4 Net Transfers 85925 6 Transfers less taxes Disposable Income 411551 2 Total income less taxes Consumable Income 364234 0 Disposable income less commodity taxes Federal Taxes 102901 3 Federal Income Tax 63575 7 UIC Contributions 8041 0 UI Benefit Recovery 64 1 CPP QPP Contributions 5857 8 O
105. sehold levels For example age and employment income are individual level variables while province and shelter expenditures are household level variables The data from the individual level are aggregated to another higher specified level for reporting purposes For example individual disposable incomes can be aggregated to census family disposable income Some of the information from the household level can be passed down for reporting at lower levels For example a table on individual capital gains by province would draw on the household level province variable All modeled variables are assigned at the individual level The level of analysis must be specified for the production of default tables user defined cross tabulations SAS files and text output files A level of analysis must also be specified for the selection facility which causes results for certain user specified units of analysis to be excluded from processing and or results The facility for displaying the univariate distributions of variables also makes use of a level of analysis The default level of analysis varies depending on the application Introduction and Overview Guide Page 20 SPSD M Version 9 2 Variable Types Variables in the SPSD M can be grouped in two broad categories Nominal or ordinal variables whose values denote membership in distinct categories are called class variables in this documentation They may also be referred to as classificatory categ
106. self uses to store data As a result data can be made accessible to SAS without the awkwardness of dealing with character file layouts appropriate variable definitions and labels etc Another benefit is that disk requirements and execution time are greatly reduced since SAS would normally have first to read in a large character file and then convert it into a large binary file before any operations could be performed on the data Many operations that can be performed with SAS such as record selection creation of new variables cross tabulation and distributional analysis can be performed using built in SPSM facilities These facilities operate many times faster than their SAS equivalents and don t require large amounts of disk space Therefore as a general rule the built in SPSM facilities should be used in preference to a SAS solution wherever possible Only if the user needs to perform a function that SPSM cannot handle should SAS be used Examples of applications that require SAS include regression analysis scatter plots and microdata analysis SPSM results file MRS The SPSM can produce a special file known as an SPSM results file This file is written using the same data compression techniques as the SPSD files Because of this SPSM results files occupy a modest amount of disk space but they can only be read by the SPSM program The results file is written using the variant results typically so that it can be read in as base result
107. sfer parameters gt y gt mprdesc Description of tax transfer parameter file 25 MPRDESC gt My 1992 Base Run 1992 SPSD M Defaults in Place 26 gt go Writing out changed tax transfer parameters to test1 mpr Changing INPVARMPR from SPSD ba92 mpr to test1 mpr 21 Default adjustment parameters are loaded from the spsd directory as specified by the default in the control parameter file 22 No changes to the standard database adjustments are required for 1992 23 The CTFLAG parameter indicates that commodity sales taxes will be calculated 24 Any desired changes to parameters that effect the execution of the variant tax transfer algorithm should be given in this phase of the dialogue 25 It is good practice to keep current descriptions in the parameter files MPRDESC is set to an appropriate label for this run 26 The go command tells the SPSM to proceed with the execution of the Introduction and Overview Guide Page 47 SPSD M Version 9 2 27 28 29 30 27 28 29 30 simulator Note that if any of the tax transfer parameters are changed a new mpr file will be written In this case the INVARMPR parameter in the test1 cpr control parameter file will be changed to the name of the output mpr file This ensures that a subsequent SPSM execution using test1 cpr will refer to the new set of tax transfer parameters Se Execution SPSM started on Fri Sep 14 1
108. sts of one or more levels separated by asterisks One of the levels the tabulation level indicates the item or items to be tabulated Any remaining levels are classification levels They specify the particular pattern of categorization desired in the table Introduction and Overview Guide Page 37 SPSD M Version 9 2 Naturally there are some restrictions on just what the X tab facility can tabulate but the design of the SPSM makes it quite flexible For starters the analyst can tabulate any analysis variable any non categorical variable available in the SPSD or calculated in the SPSM Complete lists of these variables appear in the Variable Guide As well the analyst can tabulate any of the user defined ex variables Further the analyst can define on the fly the desired tabulation expressions constructing them from any of the preceding types of variables and appropriate mathematical operators An example of a table request is EF hdprov units CF immtot units immtot units hdtenur This request would generate two tables The first would contain values for the number of economic families in each province as well as a total row The second table to the right of the semi colon would tabulate total income accruing to census families the numbers of such families and the average income per census family all broken out by tenure categories rented owned with mortgage etc as the columns DISTRIBUTIONAL ANALYSIS FACIL
109. taxes since taxable income has decreased Quebec is an exception to this because Family Allowances are excluded from net income in that tax system in 1992 The increase in the Child Tax Credit does not affect provincial treasuries since that program is non taxable The varying impact of the change by province can also be seen Introduction and Overview Guide SPSD M Version 9 2 Page 71 Table 2U Change in Consumable Income M for Census Families by Consumable income group TM Base Change in Consumable Consumable Income Group Income TM Min 5000 8 6 5001 10000 15 6 10001 15000 1 5 15001 20000 8 4 20001 25000 37 8 25001 30000 224 30001 35000 13 1 35001 40000 2 1 40001 45000 13 5 45001 Max 22 2 oM oM All 67 4 oM o Table 2U shows the flows between income groups that the scenario produces The results which show middle income families gaining at the expense of mainly poorer ones may seem at first surprising since the Child Tax Credit is by design more progressive being a refundable tax credit than Family Allowances a taxable demogrant The difference can be explained by observing that the combined Child Tax Credit and Family Allowance per child for those paying no tax was 1020 in 1992 Our example scenario on the other hand decreased this amount to 906 25 The fact that higher income families lose is not surprising since they lose Fam
110. te the household sequence numbers of the top five losers using the distributional analysis facility A second run which would select only the five households of concern and print out all variables for those households could then be submitted Aggregate results take three distinct forms built in tables whose format is fixed by SPSM user specified cross tabulations or the report produced by the distributional analysis facility Tables can be printed analyzed directly or used as input in other software packages Spreadsheet packages can be particularly useful for manipulating the results from aggregate tables Please see the Tools User s Guide for more information on the spreadsheet interface facilities supplied with the full SPSD M DISPOSABLE AND CONSUMABLE INCOME An important distinction to keep in mind when using the SPSD M is the difference between disposable and consumable income The analysis of federal and provincial taxes paid by individuals and families should include commodity and sales taxes especially as these types of taxes play a larger role in the overall tax system The tax transfer parameters supplied with the SPSD M support analyses which are concerned with both income and sales tax Disposable income is defined in the SPSD M as total income minus total federal and provincial income taxes It therefore represents the amount of income an individual or family has available for spending e g shelter food savings Consu
111. ter file also acts as documentation for a program run because it contains all the information necessary to make the run Control files normally have a cpr extension Database Adjustment Parameters Database adjustment parameters allow the values of variables on the database to be changed in order to represent dollar values in calendar years other than 1997 For example the dividend income of each individual on the database is multiplied by the database adjustment parameter GFIDIV In 1987 the value was 1 201 and thus all dividend income is 20 1 percent higher Some of the database adjustment parameters are more structural in nature For example the parameter PTF specifies Introduction and Overview Guide Page 27 SPSD M Version 9 2 poverty income thresholds for the default tables and the parameter UITARGYRMAX establishes maximum insurable earnings levels for unemployment insurance If any database adjustment parameters are modified from their initial values in the database adjustment parameter files during the user dialogue then a modified database adjustment file with file extension of apr will be generated The database adjustment parameters supplied with the full SPSD M are based on differences between SPSD per capita base year 1997 figures and per capita national accounts data for available years For years where projections must be made a simple extrapolation of the growth rate over the past two years is performed Users may provid
112. ters GFISENF 10 3t Growth Factor Self employed income non 0 74940 0 89140 0 85640 0 79350 0 87880 0 77390 0 79200 0 75870 0 79820 0 85950 GFIEMP 10 Growth Factor Employment income 0 96450 0 96450 0 99370 0 99370 0 99370 0 99370 0 96230 0 96230 0 95560 0 95560 0 94580 0 94580 0 97530 0 97530 0 88710 0 88710 0 90040 0 90040 0 95650 0 95650 12 Do you wish to modify any database adjustment parameters gt n The name of the file containing the database adjustment parameters which will be loaded has previously been specified in the control parameter file Adjustment parameters are always applied In this case i e in file spsd ba96_92 apr all growth factors have been set to 1 11 and effectively the database is not altered Because we wish to use these defaults we will respond n to the prompt 12 Lz I2l lz2l2z2l2l2l222 2 2 Tax Transfer Parameters Loading variant tax transfer parameters from SPSD ba92 mpr Current values of selected variant tax transfer parameters MPRDESC Current values for 1992 13 CTFLAG 1 Commodity tax activation flag CTOPT 2 Commodity tax calculation method CTDFLAG 0 Commodity tax detailed calculation flag 14 TARGETYEAR 1992 Year of analysis Do you wish to modify any variant tax transfer parameters gt n L l2llz2l2lzl2llz2ll22ll2 l2 2 2 222 Execution SPSM started on Mon Sep 17 08 34 53 2001 15 SP
113. that have not yet been defined nfempinc is the combined employment income of the head and the spouse of the census family which will be subsequently used for income testing This is accomplished by specifying the employment income of the individual being processed idiemp plus the employment income of the spouse SP diemp using the modifier SP to indicate the value of the variable should be read from the spouse Benefits will be restricted to individuals over the age of 20 This is accomplished by establishing a conditional group of statements with in the braces See the User s Guide for more information on statements The EITC benefit payable to eligible individuals having a combined head spouse employment income of less than 8 000 is then calculated as 1596 of head plus spouse employment income This is the first slope of the function Note that zero employment income will result in zero benefits Introduction and Overview Guide Page 75 SPSD M Version 9 2 7 The maximum EITC benefit of 1 200 is then payable to eligible individuals having a combined head spouse employment income of between 8 000 and 12 000 8 The EITC benefit payable to eligible individuals having a combined head spouse employment income of between 12 000 and 24 000 is then calculated as maximum benefits less 10 of employment income in excess of 12 000 This is the third and final slope of the function 9 The variable eitc is then labelled 10 dispinc immd
114. the total income and divide it by the employment income of that person This produces the proportion of total income accounted for by employment income Now consider the term to the left of the If an individual is over 20 years old and less than 65 years old then the first term will evaluate to 1 and the value of the expression becomes the value of the second term the employment income ratio If however the individual is under age 21 or over age 64 then the first term and consequently the entire expression will evaluate to O Note that in this example the end result if tabulated using the X tab facility becomes an average of ratios If we had wanted to examine the ratio of averages we would simply include the relevant terms of the expression in a custom table request PARAMETER EDITING Parameter editing is performed in the user dialogue during a program run or with a text editor prior to a program run User Dialogue Commands The user dialogue session is the first step in an SPSM program run It provides an interactive session for changing and examining parameter file contents As already mentioned the dialogue proceeds in three stages that correspond to the three types of parameter files Control parameters come first followed by database adjustment parameters and then by tax transfer parameters Following the last stage control of the program run is passed from the user to the simulator If the user elects to modify parameters i
115. then output in different forms for analysis The entire process is controlled by parameters Any complete simulation involves all of the following four components Database Parameters Simulation Outputs Each of these four components must be understood on its own as well as in terms of its relationship to the other components as shown broadly below SPSD Database This section describes the function and most important features of each of these four components It also describes the ways that these components interact with each other Table Outputs Microdata Outputs SPSM Parameters The SPSD database forms the foundation for the entire SPSD M and is discussed first in terms of its scope organization structure and variables The simulator performs all operations on the database including adjustment and screening simulation of modeled variables and reporting results The parameters that control this process are discussed next This is followed by a general discussion of the outputs available from the SPSM The section concludes with an examination of the SPSD M files and their relationship to the components just discussed DATABASE The SPSD was constructed by combining individual administrative data from Introduction and Overview Guide Page 17 SPSD M Version 9 2 personal income tax returns and unemployment insurance claimant histories with survey data on family incomes and on expenditure patterns The techniques
116. ther Recoveries 671 0 Federal Commodity Taxes 24691 8 Federal Transfers 69582 5 Federal Family Allow 3038 5 3 0 billion in Federal Family Allowances CTC Child Benefits 2314 1 Total Entitlement not just refundable portion OAS 13296 7 GIS 3296 4 SPA 359 8 UI Benefits 17196 0 CAP Federal Portion 4411 5 CPP QPP Income 17347 6 Other Federal Transfers 5467 5 Sales Tax Credit 2854 2 Que Tax Abatement ref 0 2 Federal Net Balance 33318 8 Federal taxes less transfers explicitly Provincial Taxes 62038 1 analyzed Provincial Income Tax 39412 7 Provincial Commodity Tax 22625 4 Provincial Transfers 9431 3 Family Programs 485 3 Elderly Programs 533 0 Tax Credits amp oth Trans 1465 5 Provincial Net Balance 52606 8 Provincial taxes less transfers lp ERE CD DOR explicitly analyzed Understanding Basic Components l CAP Provincial Portion 6947 4 T Introduction and Overview Guide Page 16 SPSD M Version 9 2 Having performed a complete SPSM run we leave aside for the moment the specifics of performing a program run and turn our attention to the several key components that form the basis of the model and any given simulation At a very general level the SPSM accepts as input a set of database variables and applies various algorithms to them which in turn produce a set of simulated variables The database and simulated variables are
117. ther calculated or read from a file We shall set it to 2 and cause base results to be calculated using the standard tax transfer algorithm Whenever BASMETH is set to 2 the INPBASMPR parameter must also have a value We shall want the base results to reflect the 1992 tax transfer system with no changes so we select the default 1992 tax transfer parameter file We can save time in the subsequent run by creating a variant results file Since consumable income is the only variable we shall be initially comparing between scenarios we will use the default which is already set to immicons in the OUTVARMRS parameter see previous page The final measure of interest to us is the change in consumable income Here the base results value is being subtracted from the variant results value to create a user defined variable called deltacon The new user variable is now appropriately labeled gt xtflag X tab facility activation flag XTFLAG gt 1 gt Xtspec X X tab specification string Introduction and Overview Guide Page 60 SPSD M Version 9 2 13 XTSPEC gt 01 CF _immicons L Base Consumable Income 02 immicons L Variant Consumable Income 03 deltacon CF _immicons L Base Consumable Income immicons L Variant Consumable Income deltacon gt go No files will be overwritten with these control parameters Any further control parameter changes gt n Writing out control parameters to factc1 cpr 14
118. thod CTDFLAG 0 Commodity tax detailed calculation flag TARGETYEAR 1992 Year of analysis 30 Do you wish to modify any variant tax transfer parameters gt y gt mprdesc Description of tax transfer parameter file 31 MPRDESC gt No federal FA Double CTC plus 10 gt ctcpc Child tax credit per child 32 CTCPC gt 1212 29 No database adjustment will be performed in any of the program runs in this tutorial session 30 Even though we read the tax transfer parameters from the FACTC1 run we Introduction and Overview Guide Page 65 SPSD M Version 9 2 must still increase the child tax credit per child 31 Good SPSM practice is again followed by keeping the file documentation up to date 32 The CTCPC parameter is increased by ten dollars to 1212 in this run 33 gt go Writing out changed tax transfer parameters to factc2 mpr Changing INPVARMPR from factc1 mpr to factc2 mpr SPSM started on Mon Sep 17 09 35 44 2001 SPSM completed on Mon Sep 17 09 35 45 2001 Elapsed time 00 00 01 Number of households processed 3990 Number of persons selected 11954 34 Variant consumable income 000 000 366016 8 Increase from base 000 000 58 3 Control parameter file factc2 cpr Table output file factc2 tbl Variant tax transfer parameter file factc2 mpr 33 This completes the setting of the parameters required for our FACTC2 run and we now proceed to the execution phase 34
119. tion of family allowances As noted before up to date documentation should always be kept in the parameter files The FAFLAG parameter controls the execution of all federal and provincial family allowances and will be set to 0 to disable those calculations This is the only change so the go command is entered to tell the SPSM to proceed with the simulation SPSM started on Mon Sep 17 08 22 27 2001 SPSM completed on Mon Sep 17 08 22 28 2001 Elapsed time 00 00 01 Number of households processed 1770 Number of persons selected 6744 Variant consumable income 000 000 146683 5 Increase from base 000 000 2015 6 Control parameter file test2 cpr Table output file test2 tbl Variant tax transfer parameter file test2 mpr The combined effect of screening records and reading a results file rather than calculating base results on execution time is significant The same run without the use of these features would take almost half again as long Introduction and Overview Guide Page 54 SPSD M Version 9 2 27 Note that the effect on consumable income of eliminating family allowances is a loss of 2 0 billion dollars This is smaller than the 3 0 billion paid out in federal family allowances as estimated in the example simulation in the Quick Start section and represents the after tax value to Canadians of the family allowances Let us now turn to examine the two user specified tables that were produced b
120. to an MS DOS prompt The user dialogue session is the first step in an SPSM program run It provides an interactive method for changing and examining parameter file contents The dialogue proceeds in three stages that correspond to the three types of parameter files Control parameters are edited first followed by database adjustment parameters and finally by tax transfer parameters Once the user dialogue is complete and a full set of parameters has been provided the user no longer interacts with the program and the simulation phase begins In this phase the database is read records are screened and tax transfer calculations are made based on the specified parameter settings When the simulation is complete SPSM generates the specified output files SPSM results are always written to files rather than being displayed directly on the screen EXAMPLE SIMULATION We are now ready to try the first simulation The complete dialogue for producing the program run is given below Inputs that the user typed have been shown in bold followed by the ENTER key Lines are numbered in square brackets for reference in the text f you are unsure whether the MS DOS environment variables are set correctly just type the following three lines in response to the MS DOS prompt where c refers to the disk drive where the SPSD M resides Introduction and Overview Guide Page 11 SPSD M Version 9 2 set PATH c spsm win32 set SPSM c spsm set SPSD c spsd
121. to another computer program on the other There are two main ways of producing printable aggregate tables either by using one of the five supplied default tables or by creating your own customized tables See the user facilities section below for further details Certain techniques useful in quantitative analysis such as regression analysis or the calculation of gini coefficients are not incorporated into the SPSM program The SPSD database files are stored in a compact binary format which is interpretable only by the SPSM programs The output facility provides users the ability to create files containing subsets of the SPSD cases or variables in several different formats for subsequent data analysis using the SPSM or other software products These formats can be read by the SPSM SAS spreadsheet programs and other analytical software Some output options also are suited to visual interpretation There are a large number of variables potentially available to the output routines The diagram below indicates the broad types of variables and their relation to the SPSM simulator The starting point is on the left with all database variables Next there are two possible sets of modeled variables one for the base case and one for the variant User variables are calculated next as a function of any of the variables created so far Finally reference variables from previous model runs may be read and processed by the output routines Build Database C
122. troduction and Overview Guide Page 64 SPSD M Version 9 2 25 The results file created with variant variables in the FACTC1 run is specified as input for base results in this run 26 The variant scenario will use the same set of parameters as was used to create consumable income in the FACTC1 MRS file We will be editing this file s contents in the third phase of the user dialogue 27 We will not be requiring a results file for this set of variant results so the OUTMRSFLAG parameter is set to 0 The OUTVARMRS and OUTMRSVARS parameter can be left as they are 28 We can now proceed to the next stage of the dialogue Loading database adjustment parameters from SPSD ba97_92 apr Current values of selected database adjustment parameters GFISENF 10 Growth Factor Self employed income non 0 74940 0 89140 0 85640 0 79350 0 87880 0 77390 0 79200 0 75870 0 79820 0 85950 GFIEMP 10 Growth Factor Employment income 0 96450 0 96450 0 99370 0 99370 0 99370 0 99370 0 96230 0 96230 0 95560 0 95560 0 94580 0 94580 0 97530 0 97530 0 88710 0 88710 0 90040 0 90040 0 95650 0 95650 29 Do you wish to modify any database adjustment parameters gt n Tax Transfer Parameters Loading variant tax transfer parameters from factc1 mpr Current values of selected variant tax transfer parameters MPRDESC _ ba92t No Federal FA CTC 2 CTFLAG 1 Commodity tax activation flag CTOPT 2 Commodity tax calculation me
123. ture of the data as well as the organization and range of variables and their sources are briefly presented here For a full discussion of the variables and database consult both the Variable Guide and the Database Creation Guide in the Reference Manual A listing and brief description of the six major types of database variables follows SPSD Variables Demographic Income Sources Unemployment Tax deductions and credits Consumption Household weights Introduction and Overview Guide Page 18 SPSD M Version 9 2 SPSD Variable Types Demographic Variables Demographic variables include age sex province and family structure A number of other classification variables such as industry occupation educational status labour force characteristics and housing tenure are also present These variables are taken from the Survey of Consumer Finance Labour Force Survey SCF of about 40 000 households Income Variables A number of variables giving individual income for persons aged 15 years or older by source are present Sources include employment income self employment income dividends interest and capital gains These variables are drawn from the SCF except for high income individuals where they are derived from personal income tax information UI EI Unemployment Insurance Employment Insurance Variables Variables providing some detail on the structure of up to two UI claims for each individual in receipt
124. uide assumes that the user is already somewhat familiar with the use of IBM type personal computers how to turn it on enter commands execute programs etc The user should know how to save copy or delete files on floppy or hard disks type file contents at the terminal and print text files Users can refer to their own operating system manual for further details The user should also be familiar with the structure and use of the MS DOS or NTFS directory system Since directories are often misunderstood by novice computer users a short explanation of directories and files is given in the section entitled Quick Start In addition some knowledge of federal and provincial tax and government cash transfer programs as well as some familiarity with the data sources is essential The SPSD M is based on several disciplines including econometrics statistics computer science mathematics and micro simulation and training in any of these areas will be an asset While the SPSD M can be used for analysis on its own other software products can greatly compliment and enhance analysis using the SPSD M Knowledge of a PC based text editor is highly recommended At Statistics Canada we use the Codewright editor other examples include BRIEF and EDIT MS DOS In a pinch Introduction and Overview Guide Page 1 SPSD M Version 9 2 some word processors can also be used for editing parameter and source code files Spreadsheets are good tools for manipulatin
125. und in the SPSM EXAMPLE directory We will use this file to change all the control parameters in the user dialogue Witha little practice the creation and use of these files becomes quick and easy Make sure that SPSMTEST is the current directory and start the SPSM program run Introduction and Overview Guide Page 51 SPSD M Version 9 2 CASPSMTEST SPSM Welcome to the Statistics Canada Social Policy Simulation Database Model SPSD M Program Version 9 0 Database Adjustment Algorithm Standard Standard Tax Transfer Algorithm Version 9 0 88 05 Alternate Tax Transfer Algorithm none aucun Compiler Operating System MSC 12 0 WIN NT Copyright c Minister of Industry 1984 2001 Press Ctrl Break to terminate En er Control Parameters 17 Enter name of input control parameter file gt spsd ba92t Loading control parameters from spsd ba92t cpr Enter specification for generating output files gt test2 Generating default output file names 17 Even though we will be using the cpi file to declare the parameters we want changed we must first specify a valid complete control parameter file The name test2 is specified to generate the names of output files Current values of selected control parameters CPRDESC Tax ransfer 1992 Population 1992 Incomes 1992 1992 5 LICENSEE Statistics Canada SPSD M licensee Selection Sample SELFLAG 0 Selection facility activation flag SA
126. used to create the database and avoid confidential data disclosure include various forms of categorical matching and stochastic imputation While no one record on the database contains information for the same individual from the four bases the database has been constructed in such a way as to provide a micro statistically representative sample of Canadians The SPSD can be used to represent the population of Canada commencing in 1988 This is done through the use of different sets of household weights that embody demographic data and projections by age sex and province for years beyond 1997 Certain adjustments have been made in SPSD to force agreement between data and known control totals For example survey weights have been adjusted to ensure that the population by age and sex corresponds to census estimates and the number of high income Canadians corresponds to the number reported by Revenue Canada Further adjustments have been made to compensate for item non response in the surveys e g to increase the number of families receiving social assistance It should be noted that the SPSD coverage does not extend to the Yukon or Northwest Territories persons residing on reservations or armed forces personnel residing in barracks The richness and complexity of the database permit comprehensive simulations but require some degree of familiarity if the data are to be processed correctly and valid conclusions drawn The fundamental family struc
127. usssss 43 Session 2 Custom Tables Screening Records User Analysis MAN ADIOS et RT t 48 Session 3 Child Benefit Trial Simulation eee 56 Session 4 New Earned Income Tax Credit Simulation 72 Guide to SPSD M Documentation 0 ccccccccccccceeessceeeseeeeeeeeeueeeueueeeueueeeueeeeeeeeees 80 SFSD M Introductory Marn dl tiotut tice tic tx ia 80 Itistallatton Olde cosiddetto beads 80 Introduction and Overview esee 80 Added ENTIER TERT TITO LP HE oes HRS HR 80 SPSD M User s Mahllal onde oras eite estne citas cibo tug aida do utis 81 RES SM RN EE 81 User X tab Facility o ee ret dud na v dude na e EM d PN DE 81 Programmer s Gd iecit pde DI po DOE eR pM 81 Commodity Tax Model User s Guide ssesseesssssssss 81 Tools User s OUO idcirco dissi desde desorden d es bntei dude bees 81 SPSD M Reference Manual sina bene arde eee ae hocce bananas 81 ANelo JAIKE Loe S oce dude c estote dasdote creo aset idea Cheadle titus 81 Parameter Guide cores Dr e a dn ERR IRS TIO E 82 Variable CUI tirhctranersrashareinnunenhianaeneh cronuarhbhountnrapaness LORD AUBL rap E ALLE 82 Database Creation Guide ccc ccccccceeccecceecceececceeeceeceeeaeeceeeaeeneeeaees 82 Introduction Welcome to the SPSD M This guide presents an introduction and overview of the product and a description of the basics required to use the model It should be
128. wever for certain types of programs the black box mode is adequate Specifically these cases are where the new program has no implications for personal taxes and is not extremely complicated The technique makes use of the user defined variables and selection facility By now you should be familiar with the SPSM dialogue and status messages so the user dialogue will not be shown in this example Instead we will show an example of how to use the SPSM in batch mode The earned income tax credit for this example has the following specifications These specifications are not meant to represent a truly realistic earned income tax credit but they are sufficiently close for the purposes of this example Eligibility Maximum Benefit Income Test Turning Points Reduction Rates Individuals aged 21 years or over are individually eligible if they reside in census families in which children under the age of 21 years are present Maximum benefits are 1 200 for each eligible person Income testing will be based on the employment income of an individual plus the employment income of their spouse if present Benefits begin with the first dollar of earned income At 8 000 dollars of employment income the benefits reach the maximum and continue at that level until 12 000 of earned income where they begin to be reduced 15 of earned income is payable as a benefit up to a maximum of 1 200 After 12 000 of earned family income maximum benef
129. xiliary executable program which can compare any of the three types of parameter files The program writes the differences to the screen in a nicely formatted table but output can be re directed to a file or to a printer if desired Introduction and Overview Guide Page 36 SPSD M Version 9 2 Built in Tables The SPSM is capable of producing ten different built in tables The tables have been designed to contain a large amount of data which can be processed to produce additional statistics So for example the tables do not provide information on average OAS benefits However they do provide information on the total value of OAS benefits as well as the number of recipients of OAS which can then be used to determine average benefits The tables are grouped into five sets of pairs The first table in each pair contains values in millions of dollars for over fifty selected variables The second table in each pair contains the number of persons families or households controlled by the user which had some non zero value for the variable reported A listing of the variables reported in these tables is shown in section 2 4 where the results of Table 0 can be seen The five groups produce the following types of tables Table 0 and 0A Canada totals Table 1 and 1A Totals by province Table 2 and 2A Totals by income group Table 3 and 3A Totals by family type Table 4 and 4A Totals by proportion of a specified income threshold The user has o
130. y factors outside the tax system This is followed by the calculation of net income The Guaranteed Income Supplement Spouses Allowances and social assistance are calculated next based on net income Federal and provincial taxes are then calculated in that order Refundable tax credits child tax credit and sales tax credit and seven provincial old age income security programs are then calculated Sales amp excise taxes are calculated next by applying Input Output based effective sales tax rates calculated with COMTAX to observed family expenditures and scaled to be consistent with simulated disposable income Reporting variables for output are calculated as the last step of the driver algorithm Simulated variables as opposed to database variables are calculated either once or twice in the simulation phase Base result variables if requested are calculated first followed by variant variables Base results if generated at all may be read from a file or calculated from the standard or alternate algorithm Variant results if generated must be calculated from the standard or alternate algorithm alternate algorithms may only be created in glass box mode Base and variant results have independently specified tax transfer algorithms and parameters Distinct sets of result variables are maintained for both the base and variant so that the user may compare the two scenarios in terms of any combination of variables at any level of analysis Us
131. y this run SPSD M Database 9 00 Mon Sep 17 08 26 00 2001 Base Description My 1992 Base Run 1992 SPSD M Defaults in Place Variant Description 1992 No Family Allowances Sample 0 0510 SELSPEC nfnkids gt 0 AGENAME Standard Table 1U Selected Quantities for Nuclear Families by Nuclear family type TM TM T To Nuclear family type Base Mean Mean Effective Effective Average Change Change Income Sales Tax FA Consumabl Disposabl Tax on FA on FA le Income e Income MM M M M M MM With Kids 1 Adult 663 572 644 19 72 With Kids 2 Adult 912 521 584 327 63 IWith Elderly 1 Adult Op o o op oO With Elderly 2 Adult oO Oo o op oO Other Adut o o op op og Other 2 Adut Oo op op o Of MM MM M M M M All 871 529 594 277 65 uM MM MM M MM M Table 1U is the name given to the first of the user specified tables An examination of this table reveals the results of the specification provided earlier see notes 13 and 14 in this example Note first that all the rows of the table have zeros except for families with children This is the result of our selection criterion The contents of the column titled Base Average FA were created from the expression deltacon units The user defined variable totalfa is the sum of federal and provincial family allowances Focussing our attention on the first row of Table 1U we see that even thoug

SPSD/M - Data Library

Contents

Download Pdf Manuals

Related Search

Related Contents