Appendix 3 - World Bank Internet Error Page AutoRedirect
Contents
1. peexp 0 037168 0 011489 0 014597 0 059739 3000 00 Command igini pcexp Output Index Gini index Sampling weight weighti Variable Estimate STD LB UB tl tt et i rt 4 1 GINI_pcexp 0 266652 0 015956 0 235305 0 297999 Now we are ready to turn to the measurement of poverty using the data from the Bangladesh Household and Expenditure Survey 1991 92 1 Compute the five main measures of poverty headcount poverty gap squared poverty gap Sen index and Sen Shorrocks Thon index for per capita expendi ture using both the food poverty line and the total poverty line derived by the cost of basic needs method in the previous exercise Food poverty line Total poverty line Headcount index Poverty gap index Squared poverty gap index Sen index Sen Shorrocks Thon index 383 APPENDIX 3 Exercises 2 Compute the headcount and poverty gap indexes for specific subgroups using the food poverty line Headcount index Poverty gap index Dhaka region Other three regions Households headed by men Households headed by women Large households gt 5 Small households lt 5 3 Repeat exercise 2 above using the total poverty line Headcount index Poverty gap index Dhaka region Other three regions Households headed by men Households headed by women Large households zy Small households lt 5 Finding and Using ado Files Ther2. Technical note Stata commands that report results also save the results so that other commands can subsequently use those results r class commands such as summarize save results in r in version 6 0 or higher After any r class com mands if you type return list Stata will list what was saved Try it APPENDIX 3 Exercises Another group e class commands such as regress save results in e and estimates list will list saved results For example e b and e V store the estimates of coefficients and the variance covariance matrix respectively There is an easier way to access coefficients and standard errors either _b varname or _coef varname contains the coefficient on varname and _se varname refers to the standard error of the coefficient 3 What percentage of people are poor by this method Other Bangladesh Dhaka regions poor using food intake method 679 4 Challenge A more sophisticated method is to regress per capita total expenditure on per capita caloric intake and then predict the expected per capita expenditure at the 2 112 Calorie level Try this regress pcexp cpcap aw weighti gen feipline _b _cons _b cpcap 2112 5 Should there be separate regression for each region Cost of Basic Needs The idea behind the cost of basic needs method is to find the value of consumption necessary to meet minimum subsistence needs Usually it involves a basket of food i
3. nificantly from earlier versions of Stata 2 A calorie is the energy required to heat one gram of water by one degree Celsius A Calorie is 1 000 calories 401 APPENDIX 3 Exercises 402 References Araar Abdelkrim and Jean Yves Duclos 2007 USER MANUAL DASP version 1 4 Universit Laval PEP CIRPEE and World Bank DASP stands for Distributive Analysis Stata Package Deaton Angus 1997 The Analysis of Household Surveys A Microeconometric Approach to Development Policy Baltimore MD Johns Hopkins University Press for the World Bank Wodon Quentin T 1997 Food Energy Intake and Cost of Basic Needs Measuring Poverty in Bangladesh Journal of Development Studies 34 2 66 101
4. In analyzing poverty is it better to use adult equivalents 3 Besides looking at the mean or the median value of consumption we can also eas ily look at the whole distribution of consumption using scatter The following plots the cumulative distribution function curve of per capita total expenditure cumul pcexp gen pcexpcdf twoway scatter pcexpcdf pcexp if pcexp lt 20000 ytitle Cumulative Distribution of pcexp xtitle Per Capita total expenditure title CDF of Per Capita Total Expenditure subtitle Exercise 1 3 saving cdf1 replace The cumul command creates a variable called pcexpcdf that is defined as the empirical cumulative distribution function cdf of pcexp in effect it sorts the data by pcexp and creates a new variable that accumulates and normalizes pcexp so that its maximum value is 1 To explore the variable try 373 APPENDIX 3 Exercises 374 list pcexp pcexpcdf in 1 10 sort pcexp list pcexp pcexpcdf in 1 10 list pcexp pcexpcdf in 10 1 Then use the code shown here to graph the cdf Feel free to experiment with the scatter command The graph is also saved in a file called cdf1 gph When you want to look at the graph later just type graph use cdf1 The cumulative distribution function curve of a welfare indicator can reveal much information about poverty and inequality For example if we know the value of a poverty line we can easily find the
5. 1 000 e Within each region communes are randomly picked where the probability that a commune is picked depends on the population of the commune in this case the commune is the primary sampling unit the psu One may survey households in a cluster within the commune for instance picking 20 households in a single village Cluster sampling is widespread because it is much cheaper than taking a simple random sample of the population Let us assume that someone has also computed a weight variable wt that represents the number of households that each representative household represents thus the weight will be small for oversampled areas and larger for undersampled areas Stata has a very useful set of commands designed to deal with data that have been collected from multistage and cluster sample surveys Information must be provided on the structure of the survey using the svyset commands Using our example we would have svyset pweight weighti strata region psu thana clear all where region is a variable that indicates the regions Having set out the structure of the survey svymean can be used to give estimates of population means and their correct standard errors and svyreg can be used to perform linear regression tak ing survey design into account Other commands include svytest to test whether a set of coefficients are statistically significantly different from zero and svylc to test linear combinations such as the
6. 4 Measures of Poverty A Simple Example In Stata open the data file example dta and browse the data using Stata Data Browser or type in the numbers shown here You should see a spreadsheet listing information exactly as presented in the following table mi Stata Editor E xi Breservel Bestoe Sot Hide pee yali EE The data consist of information on consumption by all the individuals in three 380 countries A B and C Each country has just 10 residents APPENDIX 3 Exercises 1 Summarize the consumption level for each of the three countries 2 Assuming a poverty line of 125 calculate the following poverty rates for each country Country A B C a Using the headcount index b Using the poverty gap index c Using the squared poverty gap index Hint The relevant formulas are provided in chapter 4 Try programming the results in Stata rather than doing the computations by hand or using Excel 3 Which country has the highest incidence of poverty Justify your answer Poverty Measures for Rural Bangladesh 1999 Now let s work with the per capita food expenditure and the per capita total expendi ture pc foodand pcexp inc intropov data final dta created in Exercise 1 and use cbnpline the cost of basic needs poverty line derived in Exercise 2 Technical note Although it is possible to program the calculation of different mea
7. PovcalNet 1 Assume a poverty line of 1 25 per person per day in 2005 prices Create a table that shows the headcount poverty rate for the six main regions East Asia and Pacific Europe and Central Asia Latin America and the Caribbean the Middle East and North Africa South Asia and Sub Saharan Africa for 1981 1993 and 2005 2 Repeat 1 but for a poverty line of 2 per person per day 3 Based on 1 and 2 which are the world s poorest regions And which regions have seen the biggest reduction in poverty over the past two decades 4 Pick a country Graph the evolution of its headcount poverty rate over time that is for every year available 1981 1984 1987 1990 1993 1996 1999 2002 and 2005 On the same graph show the headcount poverty rate for the region in which the country is located Relative to the region has the country you chose done relatively well or poorly in reducing poverty over time 395 APPENDIX 3 Exercises 396 Pick any two countries Compute the headcount poverty rate for each country at a dozen different poverty lines 1 00 a day 1 25 a day 1 50 a day and so on and graph these curves The horizontal axis will show the poverty line and the vertical axis will show the headcount poverty rate These are poverty incidence curves Which country has the higher poverty rate Explain Exercise 9 Chapter 11 Panel Data The goal in this exercise is to create a panel of data The Banglade
8. corresponding percentage value of people below the line Suppose the poverty line is 5 000 Then the command sum pcexpcdf if pcexp lt 5000 will give the poverty rate under the max heading For consideration Why is the mean not the appropriate measure of poverty here 4 Keep pcfood pcexp pafood paexp famsize2 hhcode merge with hh dta sort by hhcode and save as pce dta in the c intropov data directory Household Weights In most household surveys observations are selected through a random process but different observations may have different probabilities of selection Therefore we need to use weights that are equal to the inverse of the probability of being sampled A weight of w for the jth observation means roughly speaking that the jth observa tion represents w elements in the population from which the sample was drawn Omitting sampling weights in the analysis usually gives biased estimates which may be far from the true values see chapter 2 Various postsampling adjustments to the weights are sometimes necessary A household sampling weight is provided in the hh dtafile This is the right weight to use when summarizing data that relate to households However we are often interested in the individual rather than the household as the unit of analysis Consider a village with 60 households 30 households have 5 individuals each with income per capita of 2 100 while the other 30 households have 10 individua
9. differences between the means of two vari ables Repeat the exercise from Household Weights and compare the results Dhaka Chittagong Khulna Rajshahi Average household size Average per capita food expenditure Standard deviation of per capita food expenditure Average per capita total expenditure Are the new weighted averages adjusted for clustering and stratification very dif ferent from the unweighted ones APPENDIX 3 Exercises Exercise 2 Chapter 3 Poverty Lines To compare poverty measures over time it is important that the poverty line itself represent similar levels of well being over time and across groups Three methods have been used to derive poverty lines for Bangladesh direct caloric intake food energy intake and cost of basic needs The following table gives a nutritional basket in per capita terms considered minimal for the healthy survival of a typical adult in a family in rural Bangladesh Direct Caloric Intake The direct caloric intake method considers any household not meeting the nutri tional requirement of 2 112 Calories per day per person as poor For this method we need to know the quantity of every food item consumed by households and its calorie content With that information we calculate the total calorie content of the food actually consumed and derive an equivalent daily caloric intake per capita for each household The data set c intropov data co
10. example we get several measures of inequality for real per capita expendi ture clpcex1 adjusted for weights given by hhsizewt and separated into urban and rural components Another helpful program is fastgini which calculates the Gini coefficient along with jackknife standard errors For example the command fastgini rlpcexl w hhsizewt jk would generate the Gini coefficient and its standard error for real per capita expenditure rlpcex1 Let s continue using per capita total expenditure to calculate inequality measures 1 Compute the Gini coefficient the Theil index and the Atkinson index with inequality aversion parameter equal to 1 for the four regions Gini Theil Atkinson All regions Dhaka region Other three regions 2 Now repeat the above exercise using decile dispersion ratios and the share of con sumption of the poorest 25 percent Statacommand xtile is good for dividing the sample by ranking For example to calculate the consumption expenditure ratio between the richest 20 percent and the poorest 20 percent you need to iden tify those two groups xtile group y nq 5 The command xtile will generate a new variable group that splits the sample into five groups according to the ranking of y from smallest to largest that is the poorest 20 percent will have group 1 while the richest 20 percent will have group 5 Similarly to identify the poorest 25 percent you need to split the sample into
11. first creating stand alone software for measur ing poverty and inequality the DAD Distributive Analysis Analyse Distributive program they then produced DASP Distributive Analysis Stata Package version 1 4 was published in December 2007 and may be downloaded from the DASP Web site http 132 203 59 36 DASP dmodules madds14 htm DASP is an add in to Stata once the program has been downloaded every time Stata is opened it is possible to click on the User button at the top of the screen and then to click on DASP which in turn provides a set of menu driven options In addition to basic measures of poverty and inequality DASP can check for dominance decompose inequality into compo nents and generate the Lorenz curve and other graphs further details are given in the manual Araar and Duclos 2007 By way of illustration here are a couple of APPENDIX 3 Exercises commands that can be used within Stata once DASP has been downloaded the first measures the headcount index producing the standard error of the estimate of the poverty rate and lower and upper bounds of a 95 percent confidence interval while the second computes the Gini index of inequality again with a standard error and confidence interval Command ifgt pcexp alpha 0 pline 3000 Output Poverty index FGT index Sampling weight weighti Parameter alpha 0 00 Variable Estimate STD LB UB Pov line tp at a ce 4
12. for the numbers Keep the predicted output yhat and residuals resid Regress the square of the residuals on the same variables as in step 1 and save the predicted value estvar Construct a variable call it lessc that is log of food poverty line estimated log of consumption square root of estimated variance Compute the probability of poverty for each household using norm flessc Construct a variable called vull that is equal to 1 if the household has at least a 50 percent probability of being poor next year Time permitting redo the exercise on the assumption that the age of the house hold head has risen by five years and the household assets have increased by 20 percent Exercise 14 Chapter 13 Simple Impact of Thai Village Fund In this exercise you will determine the impact of the Thailand Village Fund The 2004 socioeconomic survey undertaken in Thailand included a module that asked questions about who borrowed funds from the Thailand Village Fund a program that provides 1 million baht US 25 000 per village which villagers administer in the form of loans liy Open Stata and open the data file which is called tvf dta available at http mail beaconhill org j_haughton This is a fairly large file but is only a subset of the full data from the 2004 socioeconomic survey and so cannot be used to make inferences about the effect of the program in Thailand we are using it for teach ing
13. four groups APPENDIX 3 Exercises top 20 bottom 20 top 10 bottom 10 Percentage of consumption of poorest 25 All Bangladesh Dhaka region Other regions of Bangladesh 3 Challenge Many inequality indexes can be decomposed by subgroups Decom pose the Theil index by region and comment on the results Exercise 6 Chapter 7 Describing Poverty Poverty Profiles In the previous exercises we computed poverty measures for various subgroups such as regions gender of head of household household size and so on Another way to present a poverty profile is by comparing the characteristics of the poor with those of the nonpoor Characteristics of the Poor Complete the following table where poor and nonpoor are defined by cbnp in Exercise 2 of all households of total population Average distance to paved road Average distance to nearest bank of households with electricity of households with a sanitary toilet Average household assets taka Average household land holding decimals Reminder a decimal is 0 01 of an acre Average household size of households headed by men Average schooling of head of household years Average age of head years Average head of household working hours on nonfarm activities per year poor nonpoor 393 APPENDIX 3 Exercises 394 More Poverty Comparisons across Subgroups Calculate the he
14. purposes only The questions and responses to them are fairly well labeled so you should be able to navigate your way through this data set without too much difficulty Answer the following questions based on the data in tvf dta Note the variable a30 is a weight variable and should be used when answering these questions 399 APPENDIX 3 Exercises 400 What proportion of households participated as borrowers Why reasons did people give for not participating In what proportions How large was the average loan requested Received What interest rates were charged For what purposes did people say they used the loans What was the default rate on the loans What fraction of borrowers had to borrow money from elsewhere in order to ae moan FP repay their Village Fund loan h How did the Village Fund affect households economic situation i What changes would households like to see in the Village Fund Distinguish between the responses of participants and nonparticipants Summarize the data 3 How would you evaluate the impact of the Village Fund Write a 200 word proposal This may seem like a narrow question but it is really asking you to think about how you might go about measuring the impact of any program or project Exercise 15 Chapter 13 Impact of Agricultural Extension In this exercise you will determine the impact of agricultural extension Download hh98big7bs dta This file has familiar
15. 2 for chapter 3 Food needs are as shown in APPENDIX 3 Exercises table A3 1 assume the cost of basic needs poverty line is the food poverty line times 1 3 Call the poverty lines foodline91 cbnpline91 foodline98 and cbhnpline98 Merge this information using thana and vill to create a single file with all the poverty lines Call it povlines91and98 dta pveg 3 4 pfish 8 7 Remember gen fpovline gen cbhnpline 1 3 fpovline Construct a poverty indicator 1 poor for 1991 and for 1998 and show the poverty transition matrix that is a simple table showing who was poor in both years in neither year in 1991 only or in 1998 only Exercise 11 Chapter 11 Quintile Transition Matrix In this exercise you will construct a quintile transition matrix and generate measures of chronic persistent and transient poverty using data from Bangladesh Preparatory Steps Open consume98 dta keep nh hhexpfd hhexpnfdand hhexpnfd2 rename each of these by appending 98 sort by nh and save under a new name such as rconsume98 dta Open consume91 dta keep the same variables sort by nh merge with rcon sume98 check that the merge has worked using tab _merge drop the _merge variable sort by nh and save as rconsume9198 dta If you have not already done so open hh98big7bs dta and rename each variable except nh by suffixing 98 For example rename vill vill98 This file has information on income Sort using nh and save un
16. Appendix Ean Exercises Introduction Working with household data sets requires a solid mastery of appropriate statistical and data management software such as Stata or SPSS This mastery comes from learning by doing We have found that students who work though the exercises in this appendix acquire the necessary mastery and are ready to tackle almost any chal lenge in working with household data The exercises build on one another so they should be done in the order given and each completed fully before proceeding to the next one Before beginning these exercises it is important to prepare the data as set out in appendix 2 If you are new to Stata you will want to work though appendix 2 if you once knew Stata and have forgotten the details a quick skim of Appendix 2 should suffice to bring back the fond memories Exercise 1 Chapter 2 Measuring Poverty We first need to construct the data set that will be used in the later exercises Household Characteristics Open c intropov data hh dta which consists of household level variables Answer the following questions 1 How many variables are there 2 How many observations households are there 369 APPENDIX 3 Exercises 370 3 There are four regions Household characteristics may vary by regions Fill in the following table Hint use the table command Dhaka Chittagong Khulna Rajshahi Total number of households Total number of population Aver
17. Dhaka Chittagong Khulna Rajshahi Average years of schooling Gender ratio of household that is female Working population with positive working hours Working population working on a farm 2 Are the sampled individuals very different across regions 3 We now examine some gender differences For males For females Average schooling years age 2 5 Average schooling years age lt 15 Average age Working population with positive working hours Working population working on a farm Average working hours per month Average working hours on farm per month Average working hours off farm per month 4 Are the characteristics of the sampled women very different from those of the sampled men Expenditure Open c intropov data consume dta It has household level consumption expenditure information Merge it with hh dta 371 APPENDIX 3 Exercises 1 Create three variables per capita food expenditure call it pcfood per capita nonfood expenditure call it pcnfood and per capita total expenditure call it pcexp Now let s look at the consumption patterns Average per capita expenditure pcfood pcexp By region Whole Dhaka region Chittagong region Khulna region Rajshahi region By gender of head Male headed households Female headed households By education level of head Head has some education Head has no education By house
18. a household s command over resources on average Also a 30 per cent allowance for nonfood expenditure is arbitrary 1 Create a new measure of total expenditure that includes the previously excluded irregular nonfood expenditure expnfd2 compute the three FGT poverty measures of per capita expenditure pcexp_n2 and compare the results with those based on the original definition of expenditure pcexp pcexp pcexp_n2 Headcount index Poverty gap index Squared poverty gap index The nonfood allowance can be estimated from data Two methods have been con sidered see chapter 4 The first finds the average nonfood expenditure for households whose total expenditure is equal or close to the food poverty line The nonfood expenditure for this group of households must be necessities because the households are giv ing up part of minimum food consumption to buy nonfood items 387 APPENDIX 3 Exercises The second finds the nonfood expenditure for households whose food expendi ture is equal or close to the food poverty line Because the second is more generous than the first the two are usually referred to cc as the lower and the upper allowances and the poverty lines constructed using them are called lower and upper poverty lines respectively 2 Try the following then compare the results of using the two poverty lines sum pcnfood aw weighti if pcfood lt foodline 1 1 a
19. adcount and poverty gap measures of poverty for the following sub groups using cbnp1ine to define poverty Headcount Poverty gap index index Household head has no education Household head has a primary education only Head had secondary or higher education Large land ownership gt 0 5 ha person Small land ownership or landless Large asset ownership gt 50 000 taka Small asset ownership lt 50 000 taka Combined with the poverty measures computed in Exercise 3 describe the most significant poverty patterns in Bangladesh Exercise 7 Chapter 8 Understanding the Determinants of Poverty Develop and estimate a model that explains log pcexp cbnpline using avail able data The regressors may include demographic characteristics such as gender of head and family structure access to public services such as distance to a paved road household members employment such as working hours on farm and off farm human capital such as average education of working members of the household asset positions such as land holding and so forth You need to identify potentially relevant variables and the direction of their effect Then put all those variables together and run the regression Report the result and discuss whether it matches your hypothesis If not give possible reasons gen y log pcexp cbnpline reg y age age2 workhour x1 x3 aw weighti The expression x1 x3 represents other explanatory variables t
20. age distance to paved road Average distance to nearest bank Household has electricity Household has sanitary toilet Average household assets Average household land holding Average household size 4 Are the sampled households very different across regions 5 The gender of the head of household may also be associated with different house hold characteristics Male headed Female headed households households Average household size Average years of schooling of head Average age years of head Average household assets taka Average household land holding acres CAREFUL For consideration How many decimal places should one report As a general rule do not provide spurious precision Reporting the average household size as 5 35368 gives a false impression of accuracy but reporting the size as 5 is too blunt In such cases 5 4 or 5 35 would be more appropriate and is accurate enough for almost all uses 6 Are the sampled households headed by males very different from those headed by females APPENDIX 3 Exercises Individual Characteristics Now open c intropov data ind dta This file consists of information on household members Merge this data with the household level data hh dta see appendix 2 if you need a refresher on merging and answer the following questions for individuals who are 15 years old or older 1 Regional variation
21. data from Bangladesh but we have now added a new variable called agextend that indicates whether a household was cho sen to participate in a program of agricultural extension that provides advice and sup port Note The variable is invented but the rest of the data set is real We now want to ask a basic question what was the impact of the agricultural extension program 1 First let us look at the raw numbers a Load hh98big7bs dta sort by the variable nh and save b Now load consume98v72 dta or equivalent sort by nh and merge nh using hh98big7bs c Check that the merge worked correctly by looking at the merge variable 2 Now compare income and consumption levels for households that did and did not get agricultural extension help a Hint 1 First create measures of total income per capita and total consumption per capita b Hint 2 Sort by agextend and then use the syntax by agextend sum hh or equivalent APPENDIX 3 Exercises c Specifically are households that got agricultural extension poorer Richer Larger Are they more reliant on farm income 3 Next let us assume that agricultural extension was provided randomly once other variables are held constant and then ask what effect the program had a Create dummy variables for each district thana The tab thana gen than command will do this nicely b Run a regression of per capita income or consumption or farm income on the agextend
22. der a new name such as revhh98 dta Now open hh91 dta sort by nh and merge using revhh98 dta As usual check that the two files have merged by examining _merge and then delete this variable Sort by nh and merge using rconsume9198 dta Save this file which is the file with which you will now work Note that prices in 1998 were 47 percent higher than in 1991 so before incomes or expenditures can be compared they must be adjusted for the price difference We will do this in the following exercises 397 APPENDIX 3 Exercises Income Exercises Construct a measure of household expenditure per capita for 1991 and multiply it by 1 47 to get the equivalent in 1998 prices Call it pce91in98 Use the xtile command to create quintiles for this variable and call them gex91in98 You may need to look up the xt ile command from within Stata to get the precise syntax Construct a measure of household expenditure per capita for 1998 Call it pce98 4 Use the xtile command to create quintiles for this variable and call them qex98 Construct a transition matrix using a simple tabulation to show how people moved from quintile to quintile between 1991 and 1998 Let the poverty line be 5 500 Work out the proportions of the households in the sample who are a Chronically poor that is average expenditure per capita is below the poverty line b Persistently poor that is expenditure per capita is always below th
23. e are a wealth of ado files on the Web and some of them are fairly easy to locate For example suppose one wants to compute the Sen index of poverty From within Stata type search Sen which will yield the following E stata SE 9 2 Results Wi re cde Prefs Data Graphics Statstes User Window Hep 6 8 6 B8 9 0000 h Search sen Keywords sen Search 1 Official help files Fags fxamples SJs and SBs srpl5_ CIS for rank stat percentile slopes differences amp ratios help cendif censlope censlope_iteration mata besf_bracketing mata bincdtree mata somdtransf mata u2jackpseud somersd somersd_mata if installed Q4 06 5 6 4 497 520 calculates confidence intervals for generalized thet sen median and other percentile slopes and per unit ratios of Y with respect to x help files also document supporting mata functions sg198 Computing poverty indices help poverty ff installed 3 99 pp 29 33 STB Reprints vol 8 pp 274 278 automates the estimation of a series of standard poverty measures from unit record income data search Now by double clicking on sg108 you will obtain the following page assuming 384 that your computer is connected to the Internet APPENDIX 3 Exercises Viewer 1 net stb 48 s9108 Command net stb 48 29108 a package sg108 from http www stata com stb stb4s STB 48 59108 Computing poverty indices INITHOR S STB insert by P
24. e poverty line c Transiently poor that is were poor in one of the two years but have average expenditure per capita above the poverty line d Never poor Exercise 12 Chapter 12 Basic Measurement of Vulnerability In this exercise you will calculate the basic measurement of vulnerability For this exercise the following information is available on the income of five households To complete this exercise fill in the blanks Hint Use Excel for this Probability of poverty at least 100 120 130 160 220 Probability of once in next two Poverty line SD ofincome poverty next year Vulnerability years 125 10 125 12 125 22 125 20 125 30 e Highly vulnerable 1 If probability of poverty next year is gt 0 5 e Somewhat vulnerable 2 If probability of poverty next year is gt Po but lt 0 5 e Not vulnerable 3 If probability of poverty next year is lt Po Note SD standard deviation 398 a Indicate here whether individual is highly vulnerable somewhat vulnerable or not vulnerable APPENDIX 3 Exercises Exercise 13 Chapter 12 Measuring Vulnerability in Bangladesh In this exercise you will measure the proportion of households in Bangladesh who were highly vulnerable to poverty in 1998 Complete the following steps i Use the 1998 Bangladesh data to construct and estimate a regression model where the dependent variable is the log of consumption per capita Use final dta or pce dta
25. e poverty measure many times each time using a new random sample drawn from the original one with replacement For this pur pose it is necessary to use macros and loops in Stata The following code is an exam ple it could be copied or typed into the do file editor and executed set more 1 local i 1 while i lt 100 use c intropov data final dta clear keep pcexp weighti cbnpline bsample _N SST pcexp aw weighti line 5000 drop _all set obs 1 gen sst S_6 IE ae Sa of save temp replace else append using temp save temp replace local i i 1 sum sst The code above repeats the calculation of the SST index 100 times the sum com mand provides the standard error of these 100 estimates APPENDIX 3 Exercises Exercise 5 Chapter 6 Inequality Measures Lorenz Curve The Lorenz curve can give a clear graphic interpretation of the Gini coefficient Let s make the Lorenz curve of per capita expenditure distribution for rural Bangladesh First we need to calculate the cumulative shares of per capita expenditure and population Reminder information on pcexp is in consume dta sort pcexp gen cumy sum pcexp weight gen cump sum weight quietly replace cumy cumy cumy _N quietly replace cump cump cump _N Second we need to plot the cumulative share of expenditure against the cumula tive share of population It is also helpful to have a 45 degree l
26. es Try the following sum pcexp aw weighti gen mu r sd invnorm uniform 10 gen pcexp_nl pcexp mu Here we assume that the measurement error is a random normal variable with a standard error as big as one tenth of the standard error of observed per capita expenditure Let us assume that the measurement error mu is additive to observed per capita expenditure Note that by design this error is independent of observed per capita expenditure and of any other household or community characteristics APPENDIX 3 Exercises Now recompute the headcount ratio and poverty gap ratio using this new per capita expenditure pcexp pcexp_n1 Headcount index Poverty gap index Are these measures different for the headcount index For the poverty gap index Now consider the following situation If the measurement error is correlated with a household characteristic for example if subsistence farmers usually underre port their consumption of own production will the measurement error prob lem be more or less severe Sensitivity Analysis Apart from taking standard errors into account it is also important to test the sen sitivity of poverty measures to alternative definitions of consumption aggregates and alternative ways of setting the poverty line For example some nonfood items are excluded from the expenditure aggregate on the basis that those items are irregular and do not reflect
27. fare indicator y Be careful the command is case sensitive and in this case FGT must be written in capital letters After Line the brackets must contain a number Instead of typing all three measures one could specify the a11 option or just some of the measures If sd is typed the command will also give standard errors for the estimates which is very useful in determining the size of sampling error The command above works when there is a single poverty line However some researchers prefer to compute different poverty lines for each household as a func tion of household size local price levels and the like Assume that these tailor made poverty lines are in a variable called povlines Now the appropriate command becomes FGT y vline povlines fgt0 fgtl fgt2 sd You can specify conditions range and weights with these commands For exam ple the following command calculates the headcount ratio for the Dhaka region based on a poverty line of 3 000 FGT pcexp aw weighti if region 1 line 3000 fgt0 Sen ado and SST ado calculate the Sen index and the SST index respectively The syntax follows the same format but does not compute standard errors So for example one could use Sen y line 1000 SST y line 1000 An ambitious attempt to create a suite of programs to measure poverty and inequality within Stata has been undertaken by Abdelkrim Araar and Jean Yves Duclos of Universit Laval After
28. hat you want to include don t feel confined to just three variables Note that if you want to include categorical variables you need to convert them into dummy binary variables if the ranking of categorical values does not have any meaning For example tab region gen reg APPENDIX 3 Exercises will generate four variables labeled reg1 reg2 reg3 and reg4 The variable reg1 takes on a value of 1 for Dhaka and zero otherwise and so on When using a set of such dummy variables in a regression one must be left out to serve as a refer ence area So for instance reg y age age2 workhour x1 x3 reg2 reg4 aw weighti would include dummy variables for the regions with Dhaka serving as the point of reference After the regression it is usually a good idea to plot the residuals against the fit ted values to ensure that the pattern appears sufficiently random This could be done by adding right after the regression command predict yhat xb predict e residuals scatter e yhat Exercise 8 Chapter 10 International Poverty Comparisons The World Bank estimates the extent and evolution of world poverty with the help of PovcalNet a software interface that is available on line at http iresearch world bank org PovcalNet jsp index jsp This exercise represents an exploration of world poverty using PovcalNet To answer this exercise you will need to use a browser such as Explorer and log in to
29. hilippe van Kerm GREBE University of Namur Belgium Support philippe vankerma fundp ac d After installation see help poverty INSTALLATION FILES click here to install 59108 poverty ado 9108 poverty hip ANCILLARY FILES click here to get 9108 pov do 9108 subcvse dta click here to return to the previous screen Double click again this time on click here to instal1 and the relevant ado file will be found downloaded and placed in the appropriate folder on your computer Once this has been done successfully you will get a screen like this one E Stata Viewer net install sg108 pkg Command ret install 39108 pkg package installation package nane sgii8 pkg fron http ma stata con sth stb48 checking sgiflB consistency and verifying not already installed installing into c ado plus installation complete click here to return to the previous screen This file is called poverty ado To find out more about it simply type help poverty This program generates many measures of poverty but not unfortu nately their standard errors For a sampling of the output try poverty pcexp aw weighti line 5000 all Exercise 4 Chapter 5 Poverty Indexes Checking for Robustness The robustness of poverty measures is important because if poverty measures are not accurate many conclusions about poverty comparisons between groups and over time may not be warranted Sampling Error For example
30. hold size Large house hold gt 5 Small household lt 5 By land ownership Large land ownership gt 0 5 acres person Small land ownership or landless Summarize your findings on per capita expenditure comparison 2 Now add another measure of household size which takes into account the fact that children consume less than adults Assume that a child age lt 15 will be weighted as 0 75 of an adult For instance a household consisting of a couple with one child age 7 is worth 2 75 on this adult equivalence scale instead of 3 Go back to the ind dta and create this variable call it famsize2 then merge the revised file with the household data and the consumption data files Create per adult equivalent expenditure variables let s call them pafood and paexp and repeat the exercise above 372 APPENDIX 3 Exercises Average per capita expenditure pcfood pcexp By region Whole Dhaka region Chittagong region Khulna region Rajshahi region By gender of head Male headed households Female headed households By education level of head Head has some education Head has no education By household size Large household gt 5 Small household lt 5 By land ownership Large land ownership gt 0 5 acres person Small land ownership or landless Compare your new results with those of per capita expenditure
31. individual variables such as gender age education family size and district dummy variables The coefficient on the agextend variable measures the impact of the program You will probably want to run a few regressions one for each output variable such as income per capita that is of interest c Are the effects measured in 3 b larger or smaller than in 2 4 Finally let us run a propensity score analysis The idea is first to create a propensity score that measures the probability that a household will get agri cultural extension and then to use this score to match each treated household that is a household that gets agricultural extension with an untreated house hold that is otherwise similar that is has a similar propensity score Here is how it might work a From within Stata use the search command to find pscore and attnd and download the relevant ado files This is mainly an issue of following the instructions b Estimate the propensity score equation This will look something like this pscore agextend sexhead other variables including district dummies pscore fhat1 comsup c Now make the comparison using nearest neighbor matching using attnd xxx agextend pscore fhatl comsup where xxx refers to the outcome variable for example consumption per capita that is of interest Notes 1 These commands were substantially revised in Stata version 8 and the syntax differs sig
32. ine the line of perfect equality as a point of reference Some of the following commands are not strictly necessary but they do help produce a nice graph sort pcexp gen equal cump label variable equal Line of Perfect Equality label variable cump Cumulative proportion of population label variable cumy Lorenz curve scatter cumy equal cump c 1 1 m i i title Lorenz Curve for Bangladesh clwidth medthick thin ytitle Cumulative proportion of income per capita Now repeat this exercise for Dhaka region and compare its Lorenz curve with the Lorenz curve for the whole rural area What conclusions emerge Inequality Measures for Rural Bangladesh There is a very useful program called ineqdeco ado that computes the Gini coef ficient generalized entropy family and Atkinson family of inequality measures By 391 APPENDIX 3 Exercises 392 typing search ineqdeco within Stata and following the instructions it is straight forward to load this ado file onto your computer As in Exercise 3 you can use these programs just like other Stata commands The syntax is ineqdeco y if w weight by When the by option is used this program decomposes inequality into the within group and between group components which is often very helpful Here is a more concrete example of the command at work ineqdeco rlpcexl w hhsizewt by urban98 In this
33. ladesh Dhaka Other regions poor using direct caloric intake method 58 8 Food Energy Intake The food energy intake method finds the value of per capita total consumption expenditures at which a household can be expected to fulfill its caloric requirement and determines poverty based on that expenditure Note that this expenditure auto matically includes an allowance for both food and nonfood items thus avoiding the tricky problem of determining the basic needs for those goods This method does not need price data either but as explained in chapter 3 it can also give very mis leading results A simple way to implement this method is to rank households by their per capita caloric intakes and calculate the mean expenditure for the group of households that consume approximately the stipulated per capita caloric intake requirement Pro ceed as follows 1 Merge cpcap with hh dta and calculate the average pcexp for the households whose per capita caloric intake is within 10 percent of 2 112 either above or below see code in following box 2 Call the average value feipline and identify the households for whom pcexp is less than feipline These households are considered poor based on the food energy intake method Create a variable feip that equals 1 if the household is poor and 0 otherwise sum pcexp aw weighti if cpcap lt 2112 1 1 amp cpcap gt 2112 9 gen feipline r mean gen feip pcexp lt feipline
34. ls each with income per capita of 1 200 The total population of the village is 450 Now suppose we take a 10 percent random sample of households picking three 5 person households and three 10 person households We would esti mate the mean income per capita to be 1 650 While this properly reflects the nature of households in the village it does not give information that is representative of APPENDIX 3 Exercises individuals the village has 150 people in 5 person households and 300 people in 10 person households Weighted by individuals per capita income in this village is in fact 1 500 Try the calculation Such computations can be done easily in Stata In estimating individual level parameters such as per capita expenditure we need to transform the household sample weights into individual sample weights using the following Stata commands gen weighti weight famsize table region pweight weighti c mean pcexp Stata has four types of weights fweight pweight aweight and iweight Of these frequency weights and analytic weights are most important Frequency weights weight indicate how many observations in the popula tion are represented by each observation in the sample It takes integer values Analytic weights aweight are especially useful when working with data that contain averages for example average income per capita in a household The weighting variable is proportional to the number of person
35. mp label variable cumpl Dhaka label variable cump2 other regions scatter cumpl cump2 pcexp if pcexpscatter intcumpl intcump2 pcexp if pcexp lt 20000 c 1 1 m i i title CDFs for Dhaka and other regions clwidth medthick thin 3 Does one distribution dominate another 4 If the two lines cross at least once then we may need to test for second order sto chastic dominance The poverty deficit curve is the integral of the cumulative dis tribution up to every per capita expenditure value After creating cump1 it may be obtained by gen intcumpl sum cump1 keep intcumpl pcexp save temp replace Create intcump2 for the rest of Bangladesh After combining variables and labeling them properly label variable intcumpl Dhaka label variable intcump2 Other regions scatter intcumpl intcump2 pcexp if pcexp lt 20000 c 1 1 m i i title Poverty Deficit Curves for Dhaka and other regions clwidth medthick thin 389 APPENDIX 3 Exercises 390 5 Does one distribution dominate another here Challenge Bootstrapping Standard Error for the SST Index The bootstrapping technique can be used to calculate standard errors of poverty measures and is especially helpful in cases where the standard errors are impossible to solve analytically for example with the SST index of poverty The idea is quite simple Repeat the calculation of th
36. mp pcfood gt foodline 9 gen line_u foodline r mean sum pcnfood aw weighti if pcexp lt foodline 1 1 amp pcexp gt foodline 9 gen line_1 foodline r mean Poverty line lower upper Headcount index Poverty gap index 3 Challenge Compare poverty measures when using per adult equivalence scale expenditure paeexp with those of using per capita expenditure Stochastic Dominance One may also explore the robustness of poverty comparisons by using stochastic dominance tests The first order stochastic dominance test compares the cumulative distribution functions of per capita expenditure Let s compare the cumulative dis tributions for Dhaka with those of the rest of Bangladesh 1 First generate the cumulative distribution function for Dhaka region Note You may need to use the hh dta file and merge it with the consume dta file you might also need to create weighti as the product of weight and famsize Note the double equal signs to represent the identity keep if region 1 sort pcexp Now create a running sum of the weighti 388 variable APPENDIX 3 Exercises gen cumpl sum weighti This normalizes cumpl so it varies between 0 and 1 replace cumpl cump1 cump1 _N keep cumpl pcexp save temp replace 2 Now generate the cumulative distribution cump2 for the rest of Bangladesh Keep cump2 and pcexp and append temp dta by append using te
37. nsume dta includes the quantity of 10 food items consumed Potatoes and other vegetables listed in the table are combined into one item called vegetables in the survey assume that the total per capita daily calorie provision of this combined item is 62 and the quantity is 177 grams 1 Use the quantity information from the data set and the calorie content informa tion from the above table to calculate each household s per capita caloric intake in Calories per day Hint The unit in the data set is kilograms per week and this needs to be converted into grams per day Table A3 1 Bangladesh Nutritional Basket Per capita normative daily requirements Average rural consumer Food items Calories Quantity gram price taka kilogram Rice 1 386 397 15 19 Wheat 139 40 12 81 Pulses 153 40 30 84 Milk cow 39 58 15 90 Oil mustard 180 20 58 24 Meat beef 14 12 66 39 Fish 51 48 46 02 Potatoes 26 27 8 18 Other vegetables 36 150 38 30 Sugar 82 20 30 49 Fruit 6 20 28 86 Total 2 112 832 Source Wodon 1997 93 377 APPENDIX 3 Exercises 378 2 Create a new variable cpcap to store this caloric intake variable Now identify the households for which cpcap is less than 2 112 These households are consid ered poor based on the direct caloric intake method Create a variable directp that equals 1 if the household is poor and 0 otherwise What per centage of people are poor by this method Bang
38. s over which the aver age was computed number of members of a household for instance Techni cally analytic weights are in inverse proportion to the variance of an observation that is a higher weight means that the observation was based on more informa tion and so is more reliable in the sense of having less variance Further information on weights may be obtained by typing help weight Now let s repeat some previous estimations with the newly created weights Dhaka Chittagong Khulna Rajshahi Average household size Average per capita food expenditure Average per capita total expenditure Are the weighted averages very different from unweighted ones The Effects of Clustering and Stratification If the survey under consideration has a complex sampling design the standard errors of estimates and sometimes even the means will be biased if clustering and strati fication are ignored Consider the following typical case of a multistage stratified random sample with clustering 375 APPENDIX 3 Exercises 376 First the country is divided into regions the strata and a sample size is selected for each region Note that it is perfectly legitimate to sample some regions more heavily than others indeed one would typically want to sample a sparsely popu lated heterogeneous region more heavily for example one person per 300 than a densely populated homogeneous region for example one person per
39. shi data come from a panel of households surveyed in 1991 and 1998 The relevant data are hh91 dta hh98 dta etc or hh91v7s dta and so on if one is using Stata version 7 Each house hold has a single id called nh number of household li Download the household data for 1998 and rename the variables except for nh For instance rename sexhead sexhead98 This is done so that when the data from the two surveys are merged it will still be possible to distinguish the 1998 numbers from the 1991 numbers Sort the file using nh and save it with a name like hh98newlabels dta Now open the household data file for 1991 sort it by nh and merge it with hh98newlabels dta Check that the villages are comparable for example using compare vill vill98 Use a paired t test to determine whether there was a significant change in the education level of heads of household between 1991 and 1998 Do the same for land holdings and access to toilets Repeat step 5 but use an unpaired t test Exercise 10 Chapter 11 Transition Matrix In this exercise you will create a transition matrix that shows the extent to which households moved into or out of poverty Le Open consume98 dta rename the expenditures by suffixing 98 Merge with con sume91 dta using nh to link the files Save as consume9198 dta Create poverty lines for 1991 and 1998 using the vprice91 dta and vprice98 dta files as set out in the Exercise
40. sures of poverty it is simpler to use programs that have been written by oth ers In Stata these programs are known as ado programs The basic version of Stata comes with a large library of such programs but for specialized work such as computing poverty rates it is usually necessary to install ado programs that have been provided on a diskette or obtained on the Web For computing poverty rates and their accompanying standard errors a useful program is FGT ado which is based on poverty ado written by Philippe Van Kerm the standard error calculation follows Deaton 1997 The FGT ado file should be put in your working directory or into a directory given by c ado plus f which you may need to create for this purpose Two other useful ado programs are SST ado for computing the Sen Shorrocks Thon poverty measure and Sen ado for computing the Sen index of poverty These files are available at http mail beaconhill org j_haughton Other ado programs are available on the Internet for an example and how to access them see Finding and Using ado Files below FGT ado can calculate the headcount index or FGT 0 the poverty gap index or FGT 1 and the squared poverty gap index or FGT 2 For example FGT y line 1000 fgt0 fgt1 fgt2 381 APPENDIX 3 Exercises 382 will calculate the headcount ratio the poverty gap ratio and squared poverty gap index using a poverty line of 1 000 and wel
41. tems based on nutritional requirements and consumption patterns and a reason able allowance for nonfood consumption 1 According to the basket in table A3 1 and the average rural consumer prices how much money does a household of four need each day to meet its caloric requirements 2 One way to derive the nonfood allowance is simply to assume a certain percent age of the value of minimum food consumption How much annual total expen diture does a family of four need if it is to avoid being poor assuming that nonfood expenses amount to 30 percent of food expenses 3 vprice dta gives village level price information on all 11 food items There fore we can actually calculate a food poverty line call it foodline and a total poverty line call it cbnpline for each village using the cost of basic needs 379 APPENDIX 3 Exercises method and merge this variable with pce dta Hint Here we need to sort both data sets and merge by thana vill Do this and create a variable cbnp that equals 1 for the poor and 0 for the nonpoor 4 What percentage of people are poor by this method Other Bangladesh Dhaka regions poor by cost of basic needs method 5 The percentage of people in poverty varies according to the three methods Which method do you consider to be most suitable here Why 6 Keep all imputed poverty lines and poverty indicators merge with pce dta and save the file as final dta Exercise 3 Chapter
42. the fact that poverty calculations are based on a sample of house holds rather than the population implies that calculated measures carry a margin 385 APPENDIX 3 Exercises 386 of error When the standard errors of poverty measures are large small changes in poverty may well be statistically insignificant and should not be interpreted for policy purposes As noted above FGT also computes the standard errors of its poverty measures if option sd is specified FGT y line 1000 fgt0 fgt1 sd 1 Now let s recompute the headcount index and poverty gap index for Dhaka and for the rest of the country using the total poverty line and compute the standard errors of the two measures as well Headcount index Poverty gap index Dhaka region Poverty rate Standard error of poverty rate Other three regions Poverty rate Standard error of poverty rate 2 Does the factor of standard errors change any conclusion about the poverty com parison between Dhaka and other regions Measurement Error Another reason we need to be very careful in poverty comparisons is because the data collected are measured incorrectly This could be due to recall error on the part of respondents while answering survey questions or because of enumerator error when entering the data into specific formats Let us simulate measurement error in per capita expenditure and then investigate what effect this error has on basic poverty measur
Download Pdf Manuals
Related Search