publications and work in progress
Looking within the household: Impacts for individual outcomes using aggregate household data, UCL, (preliminary and incomplete), May 2007 PDF SLIDES
While most development policies target individuals, the estimation of the program impact on individual outcomes might be a priori impossible due to lack of individual level data on key variables, such as food consumption. This paper deals with an approach to estimate the program impact on individual impacts when only aggregate household data is available. A methodology for inferring individual outcomes from household level data is described and used to estimate the impact of an intervention (PROGRESA) on individual caloric intake. Our main results show that the program is having a stronger impact at younger ages (both for males and females) and for females up to age 30 (mothers). These ndings are remarkably in line with the program's design. Another issue explored here is whether there is asymmetric information within the household as regards food consumption. Preliminary evidence show that women might not have a complete information on food intakes of adult male members within the household.
The Effect of Income on Demand for Micronutrients in Poor Rural Mexico, (with E. Skoufias, T. Gonzalez-Cossío, and S. Rodríguez Ramirez), submitted, Aug 2006 PDF
We focus on the links between a list of macro and micronutrients and income in a sample of poor rural households in Mexico in the areas where Programa de Apoyo Alimentario (PAL), a targeted nutritional program, is operating. We explore the relationship between nutrients and income using two different econometric approaches. First, we estimate the function linking nutrients and income with a semi-parametric estimator (partially linear model). We find that a linear form is a good description of the relationship between nutrient and per capita expenditure (PCE) for households around the median of PCE in our sample. Second, we estimate the nutrient-income elasticity using a linear regression model controlling both for the clustered nature of our data and for the bias due to measurement error in nutrient consumption at the household level. Our preferred estimates (instrumental variable-fixed effect specification for the sample of all households) show a sizeable positive elasticity for some nutrients (especially vitamin A 0.8, vitamin C 0.69 and calcium 0.452). For other nutrients the effect of income on the intake is still significant but very small (elasticity for fiber is only 0.091 and for iron 0.077).
Conditional Cash Transfers, Work Incentives and Poverty, (with E. Skoufias), World Bank Policy Research Working Paper 3973, Impact Evaluation Series n.5 LINK
Conditional cash transfer (CCT) programs aim to alleviate current poverty through monetary and in-kind benefits, as well as reduce future levels of poverty by encouraging investments in education, health and nutrition. The success of CCT programs at reducing current poverty depends on whether, and the extent to which, cash transfers affect adult work incentives. In this paper we examine whether the PROGRESA program of Mexico affects adult participation in the labor market, and overall adult leisure time and we link these effects to the impact of the program on poverty. Utilizing the experimental design of PROGRESA’s evaluation sample, we find that the program does not have any significant effect on adult labor force participation and leisure time. Our findings on adult work incentives are reinforced further by the result that PROGRESA leads to a substantial reduction in current poverty. The poverty reduction effects are stronger for the poverty gap, and severity of poverty measures.
Calories, Quality of Food and Income (with E. Skoufias), The World Bank mimeo, Sep 2005
This paper studies the relationship between calorie, price of calorie and expenditure as well as the behavior of food diversity as expenditure increases. We argue here that these aspects have to be studied and interpreted in a unified way in order to fully understand the extent to which nutrition responds to income and we stress particularly that the relationship between calorie demand and expenditure gives only an incomplete view of the nutrition-income relationship. This issue is studied with data from a program targeting nutrition in poor rural Mexico, Programa de Apoyo Alimentario (PAAL). Given that one of the forms of support provided by PAAL is the provision of a food basket to the targeted households, the study of the issue of food diversity will be extremely relevant in this program’s context. In addition to this, the composition of the food basket arises as an extremely importance task in the areas covered by PAAL where poor households take their calories mostly from cereals and, therefore, there is scope for fostering food consumption diversity. Our preferred estimates suggest a very high incidence of the within-cereals substitution (especially for poor households). In addition, we also find some evidence of the substitution between cereals and other food groups only explaining a small part of the total between-group substitution. These results suggest that the food basket should be composed in such a way to offer a more varied diet with less foods from the food group cereals and more from the other food groups (especially fruit and vegetables).
Using theory for measurement: an Analysis of the Behaviour of Underground Economy, (with B. Chiarini and F. Busato), WP 2005-19 Dept of Economic, University of Aarhus Denmark, under revision, Jul 2005 PDF
This paper generates high frequency data for the underground
labor and the underground production using a theoretical general equilibrium
model, over the sample 1970:01-1992:04 (32 years; 128 observations). We compare
selected time series properties of the generated series with those of the
corresponding series estimated with classical methodologies. The generated
series for underground labor and underground production present a wider range
and are more volatile than all other series estimated with classical
methodologies. The analysis, next, suggests that the underground labor is
pro-cyclical with respect to the GDP, that is lagging it by approximately one
quarter, and that underground labor series generated from the theoretical model
are highly persistent. Finally, the estimated correlation between the cyclical
component of our generated-from-theory underground labor productivity and the
actual series of aggregate GDP is negative (-0.34), while official yearly
estimates present a positive (but very low) correlation with the cyclical
component of GDP (0.12). This suggests that the underground sector has a
positive impact over the productivity at the business cycle frequency, while it
dampens productivity fluctuations at a lower frequency.
Medium-run Effects of Oportunidades on Consumption in Rural Areas, (with O. Attanasio), IFS mimeo, English version
Chapter 1 Volume IV of Oportunidades Evaluation Document 2004, Dec 2004, Spanish version PDF(spanish) PDF(english)
This paper attempts to study the short and medium run effect of Oportunidades program on consumption and its components in rural areas. To do so, it uses the new rural evaluation sample collected by Oportunidades in 2003. In the paper we perform three exercises trying to take into account the non random allocation of the program between the first two treatment groups (1998 and 2000) and the new control group (2003). First, we compare the new control with the original sample (including localities incorporated in 1998 and 2000) in terms of several observable variables. In addition, we compare consumption expenditure and its components between these 2 groups with Propensity Score Matching methods and full interacted linear matching. Results of this exercise are not satisfactory (some impact are estimated to be negative albeit not significantly different from zero) and our interpretation is that these results are largely affected by the non availability of a baseline survey that could be used to control for unobservable pre-existing differences between treatment and control communities. The other exercise we perform is to evaluate the medium run effect of the program on consumption and its components. To do this we first compare households that have been exposed 5.5 years to the programme (programme started in 1998) to household with 3 years of exposure (programme started in 2000). Significant differences are found for education and children clothing (positive) and alcohol and tobacco (negative). We also check whether we find any significant exposure effects on durables and assets. We observe positive effects only on the availability of running water and the probability of owning a car. Finally, we exploit the remarkable variability in the number and amounts of payments received by the households in the sample. This variability might be endogenous (as it is related by household choices), therefore we use an Instrumental Variable (IV) approach. For some of the outcomes of interests, notably education and children clothing, greater exposure leads to larger effects.
Evaluation of the Impact of PROGRESA on Nutrition: Theory, Econometric Methods and an Approach to deriving Individual Welfare Findings from Household Data, UCL mimeo, Sep 2004 PDF
This paper deals with the evaluation of policy interventions aiming to improve welfare. In particular, the importance of childhood nutrition in breaking the intergenerational transmission of poverty and the reasons calling for a rigorous evaluation of the interventions are analysed. The impact of an anti-poverty program in rural Mexico, PROGRESA, on the nutrition-based indicators of health is estimated with different econometric techniques. The average PROGRESA effect on the value of food consumption and on the caloric availability is estimated exploiting the randomization as identification condition and taking into account the issue of densificaciòn. Our preferred estimates show an average increase, in the value of the food consumed weekly by the household of 23,53 pesos (13,9%) as well as an increase in the daily per capita calories consumed of about 198 kilocalories (9%). The PROGRESA impact on the calorie consumption is further analysed with non parametric techniques. PROGRESA has a positive impact across the entire distribution of the caloric consumption, that is the non parametrically estimated distribution shifts to the right due to the PROGRESA. In addition, a non parametric regression approach is implemented to analyze the relationship between caloric availability and food expenditure. Our result shows the calorie intake-food expenditure elasticity being on average 0.326 below the median of the x variable (log per capita food expenditure) and 0.063 above the median. Furthermore, this dissertation deals with the issues of obtaining individual findings from household data. Our estimated (implementing a GMM approach) equivalence scale between children and adult shows a child being only 73% of an adult as far as the consumption of food is concerned. The general result is that the individual welfare measures and PROGRESA impacts are substantially positively affected by the use of our estimated equivalence scale in place of the standard un-weighted household size. Hence, our evidence supports the view that household data may be somewhat misleading for estimating individual welfare measures. Finally, a
state-of-the-art method is implemented for deriving a relationship between calorie intake and age. Our estimates show some evidence of a more marked PROGRESA impact at younger ages.
The NAIRU: Estimation and the Effect of Permanent Sectoral Employment Reallocation”,
Rivista di Politica Economica, Nov-Dec 2002, JEL codes: C32, C51, E24 PDFThis paper analyses the NAIRU making use of a cointegrated VAR and of Italian labour market data. It is shown in the paper that a cointegrated VAR represents a statistically adequate approach to the estimation of the NAIRU, that is an effective way to overcome several problems affecting standard structural approach. In particular the paper investigates whether permanent employment shift across italian industrial sector (in order to measure sectoral shifts in labour demand it is used Neumann and Topel [34]'s employment-based dispersion index) has on effect on unemployment. Findings of this paper indicate that permanent sectoral employment shifts affect unemployment both in long- and in short term. Another interesting result is that the effect of sectoral reallocation on unemployment rate occurs only after somedelay. [JEL Code: C32; C51; E24]