Documente Academic
Documente Profesional
Documente Cultură
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=astata.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
American Statistical Association is collaborating with JSTOR to digitize, preserve and extend access to Journal
of Business & Economic Statistics.
http://www.jstor.org
Estimating the Benefit Incidence of
an AntipovertyProgram by
Propensity-Score Matching
Jyotsna JALAN
IndianStatisticalInstitute,New Delhi 110016, India(jjalan@isid.ac.in)
Martin RAVALLION
The WorldBank, Washington,DC 20433 (mravallion@worldbank.org)
We apply recent advances in propensity-scorematching (PSM) to the problem of estimating the dis-
tributionof net income gains from an Argentineanworkfareprogram.PSM has a numberof attractive
featuresin this context, including the need to allow for heterogeneousimpacts, while optimally weight-
ing observed characteristicswhen forming a comparisongroup. The average direct gain to the partici-
pant is found to be about half the gross wage. Over half of the beneficiariesare in the poorest decile
nationally, and 80% are in the poorest quintile. Our PSM estimatoris reasonablyrobust to a number
of changes in methodology.
Antipoverty programs often require that participantsmust This articleestimatesthe income gains from a workfarepro-
work to obtain benefits. Such "workfare"programshave been gram, and how those gains vary with preinterventionincomes.
turned to in crises, such as due to macroeconomic or agro- To draw a statistical comparison group to workfare partici-
climatic shocks, in which a large numberof poor able-bodied pants from a larger contemporaneousand comparablesurvey
people have become unemployed. Typically, the main aim is of nonparticipants,we apply recent advances in propensity-
to raise the currentincomes of poor families hurtby the crisis. score matching (PSM) methods, following Rosenbaum and
(On the argumentsand evidence on this class of interventions, Rubin (1983). Matchingmethods have been quite widely used
see Ravallion 1991, 1999; Besley and Coate 1992; and Lipton in evaluations,but there have as yet been few economic appli-
and Ravallion 1995.) cations of matching based on the propensity score. Some
To assess the distributionalimpact of such a program,we exceptions are Heckman,Ichimura,and Todd (1997), Dehejia
need to measure the income gains to participantsconditional and Wahba(1998, 1999), Hotz, Imbens, and Mortimer(1999),
on preinterventionincome, where the income gain is the dif- Lechner (1999), and Lopez (1999).
ference between household income with the program and We study the TrabajarProgram,an antipovertyprogramof
that without it. This conditionalimpact estimate is commonly the Government of Argentina, supported by a World Bank
referredto as a program's"benefitincidence."The "with"data loan and technical assistance. A number of features of this
can be collected without much difficulty. But the "without" setting lend themselves to PSM methods. As is common in a
data are fundamentallyunobservedsince an individualcannot crisis, other evaluationmethods requiringrandomizationor a
be both a participantand a nonparticipantof the same pro- baseline (preintervention)survey were not feasible. However,
gram. This is a well-known and fundamentalproblem in all it was possible to do a postinterventionsurvey in which the
causal inferences (Holland 1986). same questionnairewas administeredto both the participants
An assumptionthat is sometimes made in benefit incidence and the nonparticipants,and in a setting in which it was plau-
analysis is that the gross wages paid are an adequate mea- sible that both groups came from the same economic environ-
sure of the income gains to participants.(See, for example, ment. The Trabajarparticipantcould be identifiedin the larger
the various assessments of the cost effectiveness of workfare survey.
programsreviewed in Subbaraoet al. 1997.) This assumption Furthermore,using kernel density estimation techniques,
would be reasonable if labor supply to a workfare program we are able to ensure that participantsare matched with the
came only from the unemployed. But that is generally not
nonparticipantsover a common region of the matching vari-
the case in practice. Moreover,even if a participatingworker ables. Any remainingbias in the matchingestimatorcan thus
were unemployed at the time she joined the program,there is be attributedto unobservedcharacteristics.The design of the
an opportunitycost of participation.Joining the programwill
program can be expected to entail considerable rationing of
leave less time for job search.There are likely to be effects on
participationaccording to observables. The sample of non-
time allocation within the household. For example, Datt and
participantsis very likely to include people who wanted to
Ravallion (1994) find that other family members took up the
displaced productiveactivities when someone joined a work-
fare program in rural India. Such behavioral responses will
reduce foregone income, althoughwe can still expect it to be ? 2003 American Statistical Association
Journal of Business & Economic Statistics
positive. Withouttaking properaccount of foregone incomes, January 2003, Vol. 21, No. 1
we cannot know the true incidence of programbenefits. DOI 10.1198/073500102288618720
19
20 Journalof Business & EconomicStatistics,January2003
participate,but were unable to do so due to, say, nonavail- incomes will overestimatethe impact on income poverty.This
ability of the program.While our applicationis well suited to holds for a broad class of poverty measures(Atkinson 1987).
above cases, bias due to unobservablescannot be ruled out. How can one estimatethe foregone income? This is a coun-
A furtheradvantageof PSM methods for this problem is terfactualconcept in that participants'incomes in the absence
that they lend themselves naturally to studying the hetero- of the programare missing data. There are several methods
geneity of programimpact. This is of obvious interest for an one might adopt to assess the counterfactual,drawingon the
antipovertyprogram,in which knowledge of the distribution literatureon impact evaluation.One can do reflexive compar-
of impacts conditional on preinterventionincomes is crucial isons by collecting baseline data on probable(eligible) partic-
to judging the program'ssuccess. ipants before the programwas instituted.These data are then
The following section discusses the evaluationproblemand compared with data on the same individuals once they have
our methods.Section 2 describesthe Trabajarprogramand our actually participatedin the program. Alternatively,potential
data. Section 3 presents the results, and offers an economic participantsare identified, and data are collected from them.
interpretation.Section 4 concludes. However, only a random subsample of these individuals is
actually allowed to participatein the program.Another pos-
sible approachis to use propensity-scorematching methods,
THEBENEFITINCIDENCE
1. ESTIMATING OF following Rosenbaum and Rubin (1983, 1985), Dehejia and
A WORKFARE
PROGRAM Wahba (1998, 1999), Heckman et al. (1997) and Heckman,
Ichimura,Smith, and Todd (1998). Here, the counterfactual
In assessing the gains from a workfare program, the
group is constructedby matchingprogramparticipantsto non-
workers' earnings are naturallythe main focus, and that will
participantsfrom a largersurvey such as the populationcensus
be the case here. However, it should be noted that earnings or an annual national budget survey. The matches are chosen
net of foregone income are only one of the potential bene- on the basis of similaritiesin observedcharacteristics.We use
fits. There could also be risk benefits from knowing that the matching methods on nonexperimentaldata to evaluate the
programexists. There may well also be benefits from the out- impact of the program.
puts, depending on (among other things) how well targeted Since most countries now have a nationally representative
the workfareprojects are to poor areas. socioeconomic survey instrument,the marginalcost of using
We first outline a simple model of self-targetingwhich pro- PSM only includes the survey of programparticipants.The
vides argumentsfor workfare,pointing to the importanceof same survey instrumentcan then be taken to a sample of par-
foregone incomes. We then describe the matchingmethod we ticipants after the programhas started,possibly with an extra
use to estimate foregone incomes. module to cover specific questions related to the program.
PSM estimates will be reliable, providedthat participantsand
controls have the same distributionsof unobservedcharacter-
1.1 The Problem istics. Failure of this condition to hold is often referred to
The following rudimentarymodel has the essential fea- as the problem of "selection bias" in econometricsor "selec-
tures necessary to characterizethe "self-targeting"argument tion on unobservables"(Heckmanand Robb 1985). Secondly,
often made in favor of workfare(Ravallion 1991). The model the supportfor the comparison and the programparticipants
assumes that foregone income from accepting a workfarejob should be equal. Finally, it is desirable that the same ques-
is F(Y), a smoothly increasing function of preintervention tionnairebe administeredto both groups, and that participants
income Y, scaled to lie between 0 and 1. Foregone income and controls are from the same economic environment.
increases with preinterventionincome due to differences in
education, experience, and so on that are naturallycorrelated 1.2 A Feasible Methodof Estimating
with both earningsand family income. The workfareprogram Benefit Incidence
offers a wage W, with F(0) < W < F(1). Workersare only
Suppose that we have data on N participantsin a workfare
concerned about the net wage gain, that is, the work alterna-
program,and anotherrandomsample of size rN(r > 1) from
tives are judged to be the same in other respects. the population.The second set of data could be the population
It is evident that,underthese assumptions,only those work- census or the national household survey that has information
ers with preinterventionincome less than F-1 (W) will partic- relevantto the participationdecisions of the individuals.Using
ipate;the programwill perfectly screen "poor"(Y < F-1 (W)) the two sets of data, we try to match the N programpartic-
from "nonpoor" (Y > F-l(W)). The schedule of gains is ipants with a comparison group of nonparticipantsfrom the
G = W -F(Y) for Y < F-I(W) and G = 0 for Y > F-I(W), population.
yielding postinterventionincomes Y + G. The two surveys must include informationthat helps pre-
In this simple model, underestimatingthe foregone income dict participationin the program.Let X be the vector of such
will lead the evaluatorto overestimatethe impact on poverty. variables.Ideally, one would match a participantwith a non-
To see why, suppose that, in assessing the gains from the pro- participantusing the entire dimension of X, that is, a match
gram, we use a biased estimate of foregone income, namely, is only declared if there are two individuals, one in each of
M(Y) < F(Y) for all Y. Then we will overestimatethe gains the two samples, for whom the value of X is identical. This
for all Y up to M-l(W). The distributionof incomes under is impractical,however,because the dimension of X could be
the biased estimate of foregone incomes must first-orderdom- very high. Rosenbaumand Rubin (1983) show that matching
inate the actualdistribution.So the errorin assessing foregone can be performedconditioning on P(X) alone ratherthan on
Jalan and Ravallion:Estimatingthe BenefitIncidenceof an AntipovertyProgram 21
X, where P(X) = Prob(D = 1IX) is the probability of par- this article, we use three differentweights, and therebyreport
ticipating conditional on X, the "propensityscore" of X. If three differentmatchingestimators.Ourfirstmatchingestima-
outcomes without the interventionare independentof partic- tor is the "nearestneighbor"estimator,where we find the clos-
ipation given X, then they are also independentof participa- est nonparticipantmatch for each participant,and the impact
tion given P(X). This is a powerful result since it reduces estimatoris a simple mean over the income differencebetween
a potentially high-dimensionalmatching problem to a single- the participantand its matched nonparticipant.(In calculating
dimensionalproblem. our mean impact, if the income of the participantis less than
The propensity score is calculated for each observationin the income of the matchednonparticipant,we treatthe impact
the participantand the comparison-groupsamples using stan- to be zero ratherthan the observed negative number.)
dard discrete choice parametric or semiparametricmodels. If the comparison sample is large enough, then "m-to-l"
Since some studies show that the impact estimatoris robustto matching with m > 1 can be used to reduce the standard
the choice of the discrete choice model, we use standardpara- errorsof comparison.However,the gain in precision achieved
metric likelihood methods to compute the propensity scores by increasing the matched comparison sample size is typi-
(Todd 1995). Propensity-scorematching (PSM) then uses the cally modest (Rosenbaumand Rubin 1985; Rubin and Thomas
estimated P(X)'s or a monotone function of it to select com- 1996). So we construct a second estimator which takes the
parison subjects. average income of the closest five matched nonparticipants,
We used the odds ratio Pi = Pi/(1 - Pi), where Pi is the and compares this to the participant's income. Following
estimated probabilityof participationfor individual i to con- Heckmanet al. (1998), we also reporta kernel-weightedesti-
struct matched pairs on the basis of how close the scores are mator where the weights are given by
across the two samples. The nearest neighbor to the ith par-
ticipant is thus defined as the nonparticipantwho minimizes Wij - Kij K (3)
[p(Xi) - p(Xj)]2 over all j in the set of nonparticipants,where j=l
in two ways. Firstly, by providing short-termwork at rel- the government's statistics office, the Instituto Nacional de
atively low wages, the program aimed to self-select unem- Estadisticay Census (INDEC), using the same questionnaire,
ployed workers from poor families. Secondly, the scheme the same interviewing teams, and at approximatelythe same
tried to locate socially useful projects in poor areas to help time.
repair and develop the local infrastructure.This article only The national survey is the Encuesta de Desarrollo Social
assesses progressagainstthe first objective (on the second, see (EDS), a large socioeconomic survey done in mid-1997. The
Ravallion 2000). EDS sample covers the populationresiding in localities with
The subprojects are proposed by local governmentaland 5,000 or more residents.The comparisongroupis constructed
nongovernmentalorganizationsthat must cover the nonwage from the EDS. According to the 1991 census, such localities
costs. The proposalshave to be viable with respect to criteria, totaled to 420 in Argentina,and represented96% of the urban
and are given priorityaccordingto ex ante assessmentsof how populationand 84% of the total population.114 localities were
well targeted they are to poor areas, what benefits they are sampled.
likely to bring to the local community,and how much the area The second datasetis a special purpose sample of Trabajar
has already received from the program.Workerscannot join participantsdone for the purposeof this evaluation.The sam-
the programunless they are recruitedto an acceptedproposal. ple design involved a number of steps. First, among all of
The process of proposing suitable subprojectsis thus key to the projects approved between April and June 1997, 300
worker participationin the program.There are other factors. projects in localities which were in the EDS sample frame
The workerscannotbe receiving unemploymentbenefits or be were randomly selected, with an additional50 projects cho-
participatingin any otheremploymentor trainingprogram.So sen for replacementpurposes. The administrativerecords on
our eligible nonparticipantpool excluded all those individu- projectparticipantsdid not include addresses,so the Ministry
als currentlyreceiving some form of unemploymentinsurance of Labor(MOL) had to obtainthese by field work. Fromthese
payments.It is unlikely that a temporaryemploymentprogram 350 projects, the Labor Ministry could find the addresses of
such as this would affect residentiallocation, althoughwork- nearly 4,500 participants.However,for variousreasons, about
ers can commute. 1,000 of these were not interviewed. The reasons given by
The wage rate is set at a maximum of $200 per month. INDEC were that the addresses were found to be outside the
This was chosen to be low enough to assure good targeting sample frame, or they were incomplete, or even nonexistent,
performance,and to help assure that workers would take up or that all household members were absent when the inter-
regular work when it became available. To help locate the viewer went to interview the household, or that they did not
Trabajarwage in the overall distributionof wages, we exam- want to respond. In all, 3,500 participanthouseholds were
ined earnings of the poorest 10% of households (ranked by surveyed. (The number of Trabajarparticipantsduring May
total income per person) in GreaterBuenos Aires (GBA) in 1997-January1998 was 65,321.)
the May 1996 PermanentHousehold Survey. For this group, We restrictthe analysis to householdswith completeincome
the average monthly earningsfor the principaljob (when this information, and those who completed all of the questions
entails at least 35 hours of work per week) in May 1996 asked of them. Also, we only consider participantswho were
was $263. This includes domestic servants.This is an unusual actually working in a Trabajarproject at the time they were
labor-marketgroup, given that they often have extra income- surveyed. Since the EDS questionnairedoes not ask income-
in-kind.If one excludes them, the figureis $334. (As expected, related questions to those below 15 or above 64 years of
the poorest decile also received the lowest average wage, and age, we also had to restrict our attention to the age group
15-64 years for our analysis. We focus on currentTrabajar
averagewages rose monotonicallywith household income per
person.) So the Trabajarwage is clearly at the low end of the participantsin the reference week, fixed at the first week of
earningsdistribution. September 1997, who received wages from the TrabajarPro-
There are other questions that the evaluation can answer. gram during August 1997. 80% of the Trabajarsample had
currentparticipantsby this definition. The remaining20% of
TrabajarI had been targetedto middle-aged heads of house-
holds (typically male). However,underthe modified program, participantsare assumedto be beneficiarieswho had left work
it was decided not to impose this restrictionsince there was by August 1, 1997 (i.e., at the startof the survey) or who had
a risk of increasing the foregone income of participantsby not yet started the Trabajarjob. With these restrictions,the
total numberof active participantsthat we have used is 2,802.
constrainingtheir ability to adjust work allocation within the
household in response to the program.In practice, however,
the restrictionson participationmay still have been imposed 3. RESULTSAND INTERPRETATION
at the local level. If that were the case on averagein all local-
3.1
Descriptive Statistics
ities, then one might expect to find that there are unexploited
income gains by increasingparticipationby the young and by In Table 1, we present selected descriptivestatistics for the
women. We will test this. Trabajarand EDS samples. The Trabajarsample has a lower
average income, higher average family size, is more likely to
have borrowed to meet their basic needs, receives less from
2.2 Data
informalsources, is more likely to participatein some form of
Two household surveys are used. One is of programpar- political organization,and less likely to own variousconsumer
ticipants, and the other is a national sample survey, used to durables(with the exception of a color TV, which appearsto
obtain the comparison group. Both surveys were done by be a necessity of life in Argentina).
Jalan and Ravallion:Estimatingthe BenefitIncidenceof an AntipovertyProgram 23
Trabajar National
sample sample
Per capita income ($/person/month) 73.205 366.596
(101.843) (792.033)
Average household size 4.894 3.448
(2.509) (1.981)
Privatepensions ($/person/month) 10.821 18.927
(36.106) (67.813)
Social pensions ($/person/month) 1.250 .749
(6.719) (6.896)
Help fromfriendsand relatives 1.515 11.893
($/person/month) (16.013) (71.977)
% of households that need to borrow 32.777 18.820
to meet basic needs (.887) (.263)
% of populationparticipatingin some 2.910 1.450
formof politicalorganization (.318) (.009)
% of households that own a telephone 22.660 66.150
(.791) (.318)
% of househouldsthat own a color TV 75.600 77.040
(.811) (.283)
% of households owning a refrigerator 26.450 48.280
with built-infreezer (.833) (.336)
% of households owning an automatic 11.660 37.680
washing machine (.606) (.326)
Male Female Male Female
Table 2 gives the percentage distributionof Trabajarpar- impact, we have only consideredparticipantswho have earned
ticipants' families across deciles formed from the EDS with at least $150 from a Trabajarjob. The minimumwage offered
households ranked by income per capita, excluding income under Trabajaris $200. Those reporting less than $150 as
from Trabajar.(The poorest decile is split in half.) This is the their Trabajarearningsmust either be in the last phase of their
type of tabulationthat is typically made in assessing such a Trabajarjob and or have misreportedtheir income. Since we
program.It assumes zero foregone income, so each participat- are interested in the impact on currently active participants
ing family's preinterventionincome is simply actual income in the program,we excluded the observationsto get a better,
minus wage earnings from the program. "albeit a more conservative"measure of the impact. We find
Table 2 suggests that a high proportion of the families that40% of the programparticipantshave a householdincome
of participantscome from poor families. In calculating the per capita that puts them in the poorest 5% of the national
population; 60% of participantsare drawn from the poorest
Table2. Locationof Trabajar in
Participants the National
Distribution 10% nationally. By most methods of measuring poverty in
of HouseholdIncome per Capita Argentina,the poverty rate is about 20%. So 75-85% of the
participantsare poor by this standard.Such targetingperfor-
Trabajar sample Nationalsample mance is very good by internationalstandards.
households Persons households Persons
Does relaxing the assumption of zero foregone income
Poorest 5% 40.2 38.8 5.0 5.6 change the results in Table 1? Using the matching methods
Next 5% 18.0 21.3 5.0 7.8 described above, we will now see whetherthat assumptionis
Decile 2 17.5 18.5 10.0 13.1
Decile 3 9.9 9.5 10.0 11.7 justified, and how much it mattersto an assessmentof average
Decile 4 6.8 5.8 10.0 10.9 gains and their incidence.
Decile 5 2.2 1.9 10.0 9.7
Decile 6 2.5 1.6 10.0 9.1
Decile 7 1.7 1.6 10.0 9.2 3.2 Propensity-ScoreMatchingEstimates
Decile 8 .6 .5 10.0 8.2
Decile 9 .4 .3 10.0 7.9 Estimating the propensity score is a crucial step in using
Decile 10 .2 .1 10.1 6.7 matching as an evaluation strategy.Different practices have
been adopted to choose a suitable specification of the partic-
Total 100.0 100.0 100.0 100.0
ipation equation (see, for example, Dehejia and Wahba 1998;
24 Journalof Business & EconomicStatistics,January2003
Heckmanet al. 1998). The underlyingprincipleis that prein- Tables 4 and 5 give our estimates of average income
terventionvariables-that are not influenced by participation gains, and their incidence according to fractiles of house-
in the program-should be includedin the regression.Existing holds ranked by preinterventionincome per capita. The lat-
applicationsto the evaluationof job trainingprogramsin the ter is not observed. To estimate it, we first estimate the gain
United States have found employmenthistories of individuals for each participatinghousehold by each of the three meth-
to be very good predictorsof participation.In our data, we do ods described in Section 2.2. We then assign each household
not have any informationon their employmenthistories prior to a decile using the same decile bounds calibratedfrom the
to the implementationof the program. However, unlike the EDS, but this time, the participantsare assigned to the decile
job-trainingprogramstudies, we have substantialinformation implied by their estimatedpreinterventionincome as given by
on community characteristicsin which the individualsreside actual income minus the estimatednet gain.
and on the householdcharacteristicsof the individuals,as well The nearest neighborestimate of the average gain is $157,
as data on their educational and demographicbackgrounds. about three-quartersof the Trabajarwage. The nearest five
We determinedthat there were over 200 potential variablesat and nonparametricestimatorgive appreciablylower gains, of
around $100. Comparingthe standarderrorsacross the three
our disposal that could reasonablybe treatedas exogenous to
estimators, we find that, as expected, there is an increase in
participation.Among these variables,many of the community
characteristicswere very importantpredictorsof participation, precision of the estimators when we move from the nearest
as were a numberof household characteristics.We estimated neighborto the nonparametricestimator.The differencein the
standarderrorsbetween the nearest neighbor and the nearest
alternativelogistic models to predict participationin the pro- five estimatoris 16%.However,the differencein the standard
gram, and selected a final model chosen on the basis of the errorsbetween the nearestfive and the nonparametricestima-
likelihood function. tor is only 7%. Thus, while there is an improvementin the
Table 3 presents the logit regression used to estimate the
precision of the estimates when we resort to "m-to-l" match-
propensityscores on the basis of which the matching is sub- ing with m > 1, the improvementis only modest comparedto
sequently done. The results accord well with expectations the increase in the numberof matches. For this reason, com-
from the simple averages in Table 1. Trabajarparticipants putational convenience and to circumvent the small sample
are clearly poorer, as indicated by their housing, neighbor- problem in the subgroupcases, the rest of the article is based
hood, schooling, and their subjective perceptions of welfare on the "nearestfive" estimate.
and expected future prospects (relative to their parents).The The averagegain using the "nearestfive" estimatorof $103
participationregressionsuggests that programparticipantsare is about half of the average Trabajarwage. Given that there
more likely to be males who are head of households and mar- is sizable foregone income, the crude incidence numbers in
ried. Participantsare likely to be longer term residentsof the Table 1 overestimatehow pro-poorthe programis since prein-
locality ratherthan migrantsfrom other areas. The model also tervention income is lower than is implied by the net gains.
predicts that (controlling for other characteristics)Trabajar Where this bias is most notable is among the poorest 5%;
participantsare more likely to be membersof political parties while the nonbehavioralincidence analysis suggests that 40%
and neighborhoodassociations. This is not surprisinggiven of participanthouseholds are in the poorest 5%, the estimate
the design of the programsince social and political connec- factoringin foregone incomes is much lower at 10%.Nonethe-
tions will no doubt influence the likelihood of being recruited less, over half of the participanthouseholds are in the poorest
into a successful subprojectproposal. However, participation decile nationally,even allowing for foregone incomes. Given
rates in political parties and local groups are still low, even that the poverty rate in Argentina is widely reckoned to be
for Trabajarparticipants(Table 1). 20%, our results suggest that four out of five Trabajarpartic-
After estimating the propensity scores for the treated and ipants are poor by Argentineanstandards.
the comparison group, we plotted them to check the com- Figure 2 gives the mean income gain at each level of prein-
mon supportcondition. We found that there were regions of terventionincome, estimatedby a locally weighted smoothed
no overlapping support. We excluded nonparticipantsin the scatterplotof the data. The mean gain falls sharply (although
not continuously) up to an income of about $200 per person
nonoverlapregion. Furthermore,since we had enough partic-
per month (which is about the median of the national distri-
ipants in each of the provinces, we tried to limit the bias due
to location differences by matching within provinces only. bution), and is roughly constantafter that. The percentagenet
Based on Table3, the mean propensityscore for the national gain is highest for the poorest, reaching 74% for the poor-
est 5%.
sample is .075 (with a standarddeviation of .125). This is, To assess the impact on poverty incidence among partici-
of course, much lower than the mean score for the Trabajar
pants,Figure 3 gives both the observedcumulativedistribution
sample, which is .405 (.266). However, there is considerable of household per-capita income and the estimated counter-
overlap in support, with only 3% of nonparticipantshaving factual (preintervention)distribution.(There is automatically
a score less than the lowest value for participants(suggest- first-orderdominance given that we have ruled out negative
ing that there is considerableunmet demand for work on the gains on a priorigrounds.)We see a 15 percentagepoint drop
scheme). Figure 1 gives the histogramsof estimated propen- in the incidence of povertydue to the programusing a poverty
sity scores for participantsand nonparticipants.After match- line of $100 per month (for which about 20% of the national
ing, the comparison group of nearest neighbors drawn from population is deemed poor). The impact rises to about a 30
the nationalsample has a mean score of .394 (.253), very close percentagepoint decline using poverty lines near the bottom
to that of the Trabajarsample. of the distribution.
Jalan and Ravallion:Estimatingthe BenefitIncidenceof an AntipovertyProgram 25
Coefficient t ratio
Cordoba 3.5084 8.395
Chaco 1.0953 2.750
La Pampa 1.2023 3.053
La Rioja 3.1152 7.505
Misiones 1.4492 3.630
Neuquen 1.0367 2.597
Salta 1.3164 3.332
San Juan 1.4462 3.513
Santa Fe 1.5063 3.897
Santiago del Estero 1.4058 3.572
Whetherhousehold is located in an emergency town -.5455 -3.284
... a settlement of 5+ years -.9622 -3.998
.. a social housing area .3536 4.479
... an area in very damaged condition -.3197 -2.747
Dwellinghas one room (besides bathroom/kitchen) .7733 7.654
... two rooms .5247 6.805
... three rooms .2734 3.902
Mainmaterialof interiorfloors is cement/bricks .3028 2.579
Wateris obtainedfrom manualpumps -.9468 -2.902
Watershortages in last 12 months -.2707 -4.535
Portablegas is used for cooking -.5661 -2.807
Householdgets hot waterthrougha centralheating service .6968 2.444
Located <3 blocks froma place where trash is placed habitually -.3360 -5.015
... <3 blocks froma place which gets flooded .2218 3.284
... in an area where there is daily collectionof trash .1795 2.016
... in an area with a water network .7348 4.396
... in an area with sewer network .2779 4.073
. . <5 blocks fromclosest publictransportation -.2674 -2.202
..<5 blocks fromclosest publicphone -.3044 -3.109
. . <5 blocks fromclosest publicprimaryschool -.4211 -4.419
... 5-9 blocks fromclosest publicprimaryschool -.3027 -3.180
... <5 blocks fromclosest neighborhoodhealthcenter .1675 2.309
... 5-9 blocks fromclosest neighborhoodhealthcenter .1678 2.315
... <5 blocks from closest pharmacy -.4265 -5.129
... <5 blocks from closest mail -.2709 -2.655
... <10 blocks froma secondary school -1.0198 -4.231
... 10-30 blocks froma secondary school -1.0127 -4.253
... 30-50 blocks froma secondary school -.4955 -1.954
... <10 blocks froma public hospital -.3943 -3.325
Safety is the majorconcern in the neighborhood .2708 2.917
It is a dangerous street for pedestriansto cross .1472 2.040
Shortages of electricity .2925 3.084
Drugaddictionproblemin neighborhood .3855 -3.786
Male 2.2307 13.961
Head of the household .3169 2.735
Spouse of the household head -.6185 -3.858
Legallymarried .2211 2.343
Separated afterbeing married .4397 2.911
Divorced .3769 2.202
Duringlast 12 months has been absent fromhousehold for >1 month -.4450 -3.182
Bornin this locality .8215 5.019
. in anotherlocalityof same province .5672 3.373
... in another province .6523 3.867
Livedhabituallyin this localityfor last 5 years .5326 4.876
Affiliatedwitha healthsystem only throughsocial work -.6388 -7.750
... with a health system through unions and private hospital -.4694 -3.839
... with a health system through social work & mutual benefit society -1.0715 -3.291
... with a health system because he is a worker -1.1213 -6.530
Currently attends an educational establishment for primary/secondary school -.7551 -2.117
Currently a student at tertiary school .8775 2.650
Dropped out of school because found syllabus uninteresting -.5386 -3.656
... he/she was finding school difficult .6700 3.048
. location of school was inconvenient -.3996 -1.951
Dropped out of school for personal reasons .3671 2.100
Taken a course in labor training in the last 3 years .4252 5.244
Never a member of a sports association .3444 2.826
Regular member of a neighborhood association with some .9705 2.482
administrative responsibilities
Regular member of a neighborhood association with no responsibilities .8259 2.526
Never a member of union/student association .5973 2.413
Member of a political party with some administrative responsibilities .7523 1.900
Member of a political party 1.6387 6.020
(continued)
26 Journalof Business & EconomicStatistics,January2003
Table3. (continued)
Coefficient t ratio
NOTE: Only significantcoefficientsin the logit regressionare reportedin the above table. For omittedcategories and for other
variablesincludedin the regression,see Addendum(availablefromthe authors).
Tables 7-9 report the net wage gains by fractiles of prein- years (typically identified as those who are new entrantsinto
tervention incomes for three different demographic groups: the job market),and workersin the age group 25-64 years.
female participants,participantsbetween the ages of 15-24 The estimates in Table 7 are not consistent with the exis-
tence of income losses due to low female participationin the
program.The net wage gains from the programaccruing to
female participantsare virtuallyidentical to the gains for male
participants.However,the distributionof female participation
of propensityscore forparticipants
is less pro-poor,as indicatedby household income per capita;
Histogram
while over half of the members of participatingfamilies are
.15
in the poorest decile nationally,this is true of less than 40%
of the members of female participants'families. This proba-
bly reflects lower wages for women in other work, making the
Trabajarwage more attractiveto the nonpoor.
.1 For the younger cohort, however, the net gains are signifi-
0 cantly higher (comparingTables 8 and 9). Foregone incomes
are lower for the young, probably reflecting their lack of
LLo experience in the labor market.Because of this, there would
be income gains from higher participation by the young.
.05 -
(To the extent that any young participantsleave school to join
the program,future incomes may suffer.) This suggests that
the older workers may well be favored in rationing Trabajar
jobs. However, the distributionof gains is more pro-poor for
"
0
0 .25 .5 .75 .9 1
Probability
of participation Table4. Net Income Gains Fromthe ProgramUsing
of propensityscorefornon-participants
DifferentEstimators
Histogram
Nearest five Nonparametric
.75 Nearest neighbor estimator estimator
Groups
Fullsample 156.770 102.627 91.678
(296.083) (247.433) (230.327)
o 125
4. CONCLUSIONS
A counterfactualincome in the absence of the programis
~ 100
-o I missing data, and assumptionswill have to be made to make
0._ up for these missing data. The assumptionsmade in program
evaluationsare often dictatedby data availability.In assessing
0• 75
the gains from antipovertyprograms-programs that are often
E
0
set up rapidly in response to a crisis-it is common to only
500
have access to a single cross-sectional survey done after the
50I
programis introduced.Propensity-scorematching methods of
I I I I
100 150 200 250 300 350 400
Pre-intervention
incomeper capita evaluationcombine a single cross-sectionalsurvey of program
participantswith a comparable larger cross-sectional survey
Figure 2. Mean Income Gain Plotted Against Preintervention Income. from which a comparison group is chosen. With sufficiently
28 Journalof Business & EconomicStatistics,January2003
detailed cross-sectionaldata on both participantsand nonpar- We find thatprogramparticipantsare more likely to be poor
ticipants,these methodscan allow an assessmentof behavioral than nonparticipantsby a varietyof both objective and subjec-
responses without preinterventionbaseline data or randomiza- tive indicators.The participantstend to be less well educated,
tion. The accuracyof this methodwill dependon how well one they tend to live in poorer neighborhoods,and they tend to
can assure that treatmentand comparisongroups come from be members of neighborhoodassociations and political par-
the same economic environmentand were given the same sur- ties. The relatively low wage rate clearly makes the program
vey instrument.The method cannot rule out the possibility of unattractiveto the nonpoor.
selection bias due to unobserveddifferences between partic- Using our model of programparticipationto find the best
ipants and even a well-matched comparison group, although matches from the national sample for each Trabajarworker,
there is evidence this may well be an overrated problem we have estimated the net income gain from the program.
(Heckmanet al. 1998; Dehejia and Wahba 1998, 1999). We find that ignoring foregone incomes greatly overstates
We have applied recent advances in matching methods to the average gains from the program, although sizable gains
Argentina's TrabajarProgram.While neither a baseline sur- of about half the gross wage are still found. Even allowing
vey nor randomizationwere feasible options in this case, the for foregone incomes, the program's benefit incidence is
programis well suited to matchingmethods. decidedly pro-poor,reflecting the self-targetingfeature of the
NOTE: These numberscorrespondto the nearestfive estimatorforthe subgroupof female participants.Standarderrorsin parentheses.
Jalan and Ravallion:Estimatingthe BenefitIncidenceof an AntipovertyProgram 29
NOTE: These numberscorrespondto the nearestfive estimatorforthe subgroupof 15-24 year participants.Standarderrorsin parentheses.
programs' design. Average gains are very similar between standarderrorsis even smaller when we compare the nearest
men and women, but are higher for younger workers.Higher five estimator to the nonparametricestimator which uses all
female participationwould not enhance averageincome gains, of the informationin the nonparticipantsample.
and the distributionof the gains would worsen. Higher par-
ticipation by the young would raise average gains, but would
also worsen the distribution. ACKNOWLEDGMENTS
We do not have to drop any participantsfrom our sam- The work reported in this article is one element of the
ple in computing the impact estimator.This is probably the ex-post evaluation of the World Bank's Social Protection II
result of having a large comparisongroup sample from which Project in Argentina. The support of the Bank's Research
we could draw our matches. Finally, as discussed in the arti- Committee (under RPO 681-39) is gratefully acknowledged.
cles by Rosenbaumand Rubin (1985) and Rubin and Thomas The article draws on data provided by the SIEMPRO unit
(1996), we find that the precision of the matchingimpact esti- of the Ministry of Social Development, Government of
mators does improve when we use "m-to-i" matching with Argentina. The authors are especially grateful to Joon Hee
m > 1. However, the increase in the precision is moderate. Bang and Liliana Danilovich of SIEMPRO for their help
Thus, matching each participantto five nonparticipantsdoes with the data. The authors' thanks also go to staff of the
not reduce the standarderrors fivefold. The reduction in the TrabajarProject Office in the Ministry of Labor,Government
NOTE: These numberscorrespondto the nearestfive estimatorforthe subgroupof 25-64 year participants.Standarderrorsin parentheses.
30 Journalof Business & Economic Statistics,January2003
of Argentina,who provided the necessary data on their pro- Heckman, J., and Robb, R. (1985), "AlternativeMethods of Evaluatingthe
gram, and gave this evaluation their full support.They also Impact of Interventions:An Overview," Journal of Econometrics, 30,
239-267.
thankthe journal's associate editor and an anonymousreferee Holland, P. W. (1986), "Statistics and Causal Inference,"Journal of the
for useful comments. Petra Todd kindly advised the authors AmericanStatisticalAssociation, 81, 945-960.
on matching methods. Useful comments were received from Hotz, J., Imbens, G., and Mortimer,J. (1999), "Predictingthe Efficacy of
FutureTrainingProgramsUsing PastExperiences,"NBER TechnicalWork-
Polly Jones, Dominique van de Walle, and seminar partici- ing PaperT0238, Cambridge,MA.
pants at the WorldBank, the IndianStatisticalInstitute,Delhi, Lechner,M. (1999), "Earningsand EmploymentEffects of ContinuousOff-
and the Instituteof Fiscal Studies, London. the-Job Training in East Germany,"Journal of Business and Economic
Statistics, 17, 74-90.
Lipton, M., and Ravallion, M. (1995), "Povertyand Policy,"in Handbookof
[Received October 1999. Revised June 2001.] DevelopmentEconomics, Volume3, eds. J. Behrmanand T N. Srinivasan,
Amsterdam:North-Holland.
Lopez, M. (1999), "Does Bilingual EducationAffect EducationalAttainment
and Labor Market Outcomes?,"Working Paper, University of Maryland
REFERENCES School of Public Affairs.
Manski, C., and Lerman,S. (1977), "The Estimationof Choice Probabilities
Atkinson, A. (1987), "On the Measurementof Poverty,"Econometrica, 55, From Choice-Based Samples,"Econometrica,45, 1977-1988.
749-764. Ravallion,M. (1991), "Reachingthe RuralPoor ThroughPublic Employment:
Besley, T., and Coate, S. (1992), "Workfarevs. Welfare:IncentiveArguments Arguments,Evidence and Lessons FromSouth Asia," WorldBankResearch
for Work Requirementsin Poverty Alleviation Programs,"American Eco- Observer,6, 153-175.
nomic Review, 82, 249-261. (1999), "AppraisingWorkfare,"WorldBank Research Observer, 14,
Datt, G., and Ravallion, M. (1994), "TransferBenefits From Public Works 31-48.
Employment: Evidence From Rural India," Economic Journal, 104, (2000), "MonitoringTargetingPerformancewhen DecentralizedAllo-
1346-1369. cations to the Poor Area are Unobserved,"WorldBank Economic Review,
Dehejia, R. H., and Wahba, S. (1998), "PropensityScore Matching Meth- 14, 331-345.
ods for Non-ExperimentalCausal Studies,"NBER Working Paper 6829, Rosenbaum, P., and Rubin, D. (1983), "The CentralRole of the Propensity
Cambridge,MA. Score in ObservationalStudies for Causal Effects,"Biometrika,70, 41-55.
(1999), "CausalEffects in Non-ExperimentalStudies: Re-Evaluating (1985), "Constructinga Control Group Using MultivariateMatched
the Evaluationof TrainingPrograms,"Journal of the American Statistical SamplingMethodsthat Incorporatethe PropensityScore,"AmericanStatis-
Association, 94, 1053-1062. tician, 39, 35-39.
Friedlander,D., Greenberg,D., and Robins, P. (1997), "EvaluatingGovern- Rubin, D., and Thomas, N. (1996), "MatchingUsing Estimated Propensity
ment TrainingProgramsfor the Economically Disadvantaged,"Journal of Scores: Relating Theory to Practice,"Biometrics,52, 249-264.
Economic Literature,35, 1809-1855. Silverman,B.W. (1986), Density Estimationfor Statistics and Data Analysis,
Heckman,J., Ichimura,H., and Todd, P. (1997), "Matchingas an Economet- London:Chapmanand Hall.
ric EvaluationEstimator:Evidence From Evaluatinga Job Training Pro- Subbarao, K., Bonnerjee, A., Braithwaite,J., Carvalho, S., Ezemenari, K.,
gramme,"Review of Economic Studies, 64, 605-654. Graham, C., and Thompson, A. (1997), Safety Net Programs and
Heckman, J., Ichimura,H., Smith, J., and Todd, P. (1998), "Characterizing PovertyReduction:Lessons From Cross-CountryExperience,WorldBank,
Selection Bias Using ExperimentalData,"Econometrica,66, 1017-1099. Washington,DC.
(1996), "NonparametricCharacterizationof Selection Bias Using Todd, P. (1995), "Matchingand Local LinearRegressionApproachesto Solv-
ExperimentalData: A Study of Adult Males in JTPA. Part II, Theory and ing the Evaluation Problem With a SemiparametricPropensity Score,"
Methods and Monte-CarloEvidence,"mimeo, Universityof Chicago. mimeo, University of Chicago.