Abstracts

Parallel Sessions B1 to B5

Full abstracts for the five contributed sessions of block B. Each session lists three papers with authors and affiliations.

Zoom links were available only on the conference day. Back to the full program.

Block A Block B Block C Block D Block E Block F Full program

B1: Elections

Elections and Inefficiency in the Allocation of Street Pavement Contracts in Mexico

Daniel Prudencio, Rice University

This paper studies the allocation of street pavement contracts in Mexico. By law, the government should conduct a public auction for the pavement of streets, nevertheless, exceptions in the law allow the government to bypass the public auction procedure and to either choose a firm of its preferences or invite specific firms for an auction with three or more firms. In this paper, I test whether the government chooses the most efficient firms when bypassing the public procurement auction, and second, I study whether the allocation of contracts follows an electoral cycle. I find that firms selected for settings with less competition are more experienced and have lower costs for complex projects. Nevertheless, for smaller projects, they seem to be less cost efficient than firms that only participate in public auctions. Counter intuitively, contracts by direct allocation or invitation to an auction are usually small. When comparing mechanisms for projects that are more complex, firms that participate in an auction by invitation seem to bid lower than through a public auction. Finally, I find a political cycle in the type of allocation mechanisms chosen by the government, but the evidence is weaker when analyzing the influence of political factors on the number of projects.

Evo Morales and Electoral Fraud in Bolivia: A Natural Experiment

Diego Escobari, University of Texas Rio Grande Valley
Gary Hoover, University of Oklahoma

This paper exploits a natural experiment based on a disruption in the official preliminary vote counting system to identify and estimate the size of electoral fraud in the 2019 Bolivian presidential elections. The results show evidence of a statistically significant electoral case of fraud that increased the votes of the incumbent Movimiento al Socialismo and decreased the votes of the runner up Comunidad Ciudadana. We estimate that the extent of the fraud is at least 2.67% of the valid votes, sufficient to change the outcome of the election. We validate our results with pseudo outcomes on a placebo analysis. We also allow for heterogeneous voting preferences across different geographical regions (e.g., rural vs. urban) and document how fraudulent polling stations were hidden in the final reporting of the votes.

Measurement of Reelection Perspectives in Bolivia Related to the Patterns of Quality of Life

Lorena Heller Vilela, UPB

The purpose of this document is to analyze the relationship between citizens’ quality of life and their decision to re elect the ruling parties in Bolivia. For this purpose, a panel data is used at municipal and departmental level from 1999 to 2018, taking into account macroeconomic behavior and a Quality of Life Indicator. This research follows the methodological strategy proposed by Ward (2015) with the inclusion of an indicator to estimate the well being of the population. Therefore, a Quality of Life Indicator is developed following Cortes et al. (1999) who used the principal components method to summarize the information of the different components that determine the level of well being of a citizen. The results demonstrate that electoral decisions are not only associated to macroeconomic state but also to the general welfare of citizens.

B2: Poverty

Endogenous Weights and Multidimensional Poverty: A Cautionary Tale

Indranil Dutta, University of Manchester
Ricardo Nogales, University of Oxford
Gastón Yalonetzky, University of Leeds

Composite measures such as the multidimensional poverty indices depend crucially on the weights assigned to the different dimensions and their indicators. A recent strand of the literature uses endogenous weights, which are determined by the data at hand, to compute poverty scores. Notwithstanding their merits, we demonstrate both analytically and empirically, how a broad class of endogenous weights violates key properties of multidimensional poverty indices such as monotonicity and subgroup consistency. Without these properties anti poverty policy targeting and assessments are bound to be seriously compromised. Using real life data from Ecuador and Uganda we show that these violations are widespread. Hence, one should be extremely careful when using endogenous weights in measuring poverty. Our results naturally extend to other welfare measures based on binary indicators, like the widely studied asset indices.

Endogenously persistent shocks and poverty traps

Mateo Arbeláez, University of Illinois at Urbana Champaign
Leopoldo Fergusson, Universidad de los Andes
Andrés Zambrano, Universidad de los Andes

Adverse shocks are common in developing countries. Since markets are incomplete, households must significantly distort their decisions to cope with these shocks. These distortions not only affect current utility but can also increase the vulnerability to future shocks, which in turn could trap households in a low income equilibrium. This paper uses Colombian longitudinal data to estimate the persistence of adverse shocks, which is found to be between 8 and 11 percentage points, and find that such persistence is mostly explained by decreases in consumption after a realized adverse shock. We also show that, while households in the lowest and highest quintiles use assets to smooth consumption, households in the middle of the wealth distribution decrease consumption significantly after a shock. We then propose a model and calibrate it with the persistence estimates to rationalize the observed behavior, and find that a poverty trap exists for the first two quintiles Informality and Poverty in Bolivia: The Trade Boom and Its Aftermath

Endogenously persistent shocks and poverty traps

Rodrigo Gonzáles Zuazo, CIESS Econométrica

This paper studies for the first time the two way simultaneous relation between informality and poverty in Bolivia. A country which possesses the largest informal output and employment in Latin America and at the same time one of the highest poverty headcount ratios in the region. Though simultaneous maximum likelihood probit models, I find that monetary poverty increases the probability of being informal by 26 percent and that working informally is irrelevant in the determination of household poverty. This illustrates that for some workers, informality is not a voluntary decision because being surrounded by poverty at home could be forcing individuals to insert into the labor market as informal workers. However, given the heterogeneity of labor informality in terms of earnings and quality of jobs, the informal condition does not change the probability for a worker to live in poverty. To explore to what extent informal earnings are helping individuals not to fall into poverty, in year Shapley decompositions are presented. I use cross sectional Bolivian microdata for two years of the trade boom period (2007 and 2012) and two years for the ex post period (2015 and 2018), where, a very stable relationship between informality and poverty is evidenced. The findings suggest that different macroeconomic and trade contexts are almost irrelevant in shaping this relationship. Finally, the paper highlights the importance of informal employment as a key factor to reduce poverty in Bolivia and challenges those visions in line with its eradication.

B3: Public Goods

Provision Effects of Local Public Goods on Crime and Education: Evidence from Colombia

Carolina Vélez Ospina, Universidad del Rosario

The provision effects of local public goods on crime and education are not clear in the literature. While some argue that provision does not affect these outcomes, others find that effects depend on the benefits it offers to the community. This paper studies the effect of the construction of cultural centers in Medellín, Colombia on crime and test scores in mathematics and language. This policy is interesting since the communities participated in the design of these cultural centers. Using a dynamic difference in differences strategy, I find that schools near centers improve their test performance, especially for younger children. Regarding crime, I find that in neighborhoods near centers, there is a reduction in motorcycle and car theft crimes.

Public good or public bad? Indigenous institutions, nation building, and the demand for road infrastructure in Mexico

Aldo Elizalde, University of Glasgow
Eduardo Hidalgo, Center for Macroeconomic Research, University of Cologne
Nayeli Salgado, Vienna University of Economics and Business (WU)

Large scale infrastructure projects are understood to be a significant threat to the survival of Indigenous people’s culture and existence. This paper investigates which Indigenous groups are better able to oppose public ‘megaprojects’, using Mexico as a laboratory. In the aftermath of the Mexican Revolution (1920s 1950s), the new state started to build a network of roads as a tool for nation building. Yet Indigenous groups saw roads as a threat to their ethnic identity and local practices. We combine newly digitised and georeferenced municipality level data on road infrastructure with detailed Indigenous ethnicity data from the early 20th century. Exploiting the expansion of roads in a diff in diff setting, we find lower road infrastructure where Indigenous people descend from centralised societies in pre colonial times. Plausible mechanisms driving our results are better coordination for collective actions and higher preferences for preserving Indigenous identity. We also show that where road building was more successful, present day economic outcomes are better.

Vehicle Stock and Ownership Elasticity for Bolivia 2000–2035

Javier Aliaga Lordemann, ABCE and INESAD
Alejandra Terán Orsini, INESAD

This paper seeks to analyze the evolution of the Bolivian vehicle stock in the mid term and their policy implications. First, we analyze the relationship between income and vehicle ownership in the country during the period 1970–2017 by robust econometric techniques. Based on these results, we used an energy mix accounting model programmed in General Algebric Model Sistem (GAMS) to analyse how the vehicle fleet evolve over the time and the derived demand of gasoline, natural gas and diesel oil. Finally, we observe the trajectory of CO2eq in the transport sector by different types of vehicle categories. Our results prove that the relationship between vehicle ownership and per capita income is highly non linear and we verify an excessive increase of vehicle fleet during the last decade, both results will speed up the saturation level of the vehicle fleet in Bolivia.

B4: Public Health and Poverty

Feminización de la pobreza en Bolivia, un análisis a partir de paneles sintéticos

Cecilia Castro Balderrama, IISEC – Universidad Católica Boliviana
Guillermo Gómez Aliaga, IISEC – Universidad Católica Boliviana

In recent years, Bolivia has been characterized by decreasing poverty rates among different groups of population. For a better understanding of this phenomenon, it is important to verify if there is any difference in relation to poverty, between men and women. Therefore, this research unites the concepts of poverty and gender to investigate the feminization of poverty in Bolivia. For this purpose, nonparametric synthetic panels are constructed to verify how the dynamics of poverty vary between households headed by men and households headed by women on 2013 to 2017 period. Results indicate that women had a pattern which was similar to the general population trend, the majority of households headed by women remained out of poverty, a smaller proportion was transiently poor, and that a minimum percentage of households was chronically poor.

Service Delivery Point and Individual Characteristics Associated with Modern Contraceptive Adoption: A Multi Country Longitudinal Analysis

Philip Anglewicz, Johns Hopkins Bloomberg School of Public Health
Carolina Cardona, Johns Hopkins University
Titilope Akinlose, Research Triangle Institute International
Amy Tsui, Johns Hopkins Bloomberg School of Public Health

Background: Women who start using contraception (“adopters”) are a key population for family planning goals, but little is known about characteristics that predict the adoption of contraception and not just its use. We used prospective data from five countries, (Democratic Republic of Congo, India, Kenya, Nigeria, and Burkina Faso) and identified baseline characteristics that predicted adoption of modern contraception in the short term. Methods: We used data from the Performance Monitoring for Action (PMA) Agile Project, which operated in urban sites of these five countries. PMA Agile administered a survey to facility managers, followed by a client exit interview survey (CEI). Female clients responding to the CEI were asked for phone numbers that were used for a phone follow up survey approximately four months later. For our analysis, we used data from the CEI baseline and phone follow up survey to compare women who start using contraception during this period with those who remain non users. We used characteristics of the facility and the woman at baseline to predict her contraception adoption in the future. Results: Discussing FP with a partner at baseline was associated with greater odds of adoption in DRC (OR 2.123; 95% CI 0.898–5.017), India (OR 2.486; 95% CI 1.163–5.315), Kenya (OR 1.653; 95% CI 1.159–2.356), and Burkina Faso (OR 1.599; 95% CI 0.963–2.654). Nigerian women who discussed family planning with any staff member at the health facility had 1.89 times greater odds (95% CI 1.231–2.885) of becoming an adopter. Women with secondary education in DRC (OR 2.666 95% CI 1.014–7.012) and Burkina Faso (OR 1.682; 95% CIs 1.006–2.812); and women with college/university education in India (OR 3.237; 95% CIs 1.129–9.276) and Burkina Faso (OR 1.907; 95% CIs 0.971–2.812) had higher odds of adopting modern contraception than women with no education. Conclusions: Characteristics of both the woman and the health facility are associated with adoption of modern contraception in the future. Some characteristics, like discussing family planning with a spouse, education, and parity, are associated with contraceptive adoption across settings. Other characteristics that predict contraceptive use, such as health facility measures, vary across countries.”

Under five mortality in Bolivia: Survival analysis and risk factors from 2003, 2008 and 2016 Health and Demographic Surveys

Pamela Córdova, UPB
Soraya Román, UPB
Zarela Soria Galvarro, UPB

Infant mortality under five years (5q0) is a factor associated with the well being of the population and an indicator of health development and the socioeconomic status of a country. According to the Health Equity Report 2016, the countries with the highest under five mortality rate in Latin America and the Caribbean are Bolivia and Haiti. This study seeks to identify the effects of individual, household and community (contextual) factors on the under five mortality rate. Initially, we estimate survival times for 2003, 2008 and 2016 health and demographic surveys using the Kaplan Meier function. Then, we use Cox’s proportional hazards model and Cox’s hierarchical (or multilevel) proportional model to compare children under five’s risk of death for different years and covariates. We find an increase in the concentration of deaths in the first year of life over time, especially during the first month of life (neonatal mortality). In addition, the study shows important differences in the risk of death respect to the region of residence (East and West for data from the ENDSA 2003 and Altiplano, Valleys and Llanos for the ENDSA 2008 and EDSA 2016). Hence, for policy recommendations, future studies should focus on the identification of 5q0 risk factors for each region of Bolivia separately.

B5: Energy

Estimaciones del PIB per cápita y de la actividad económica a nivel municipal en Bolivia en base a datos de consumo de electricidad

Lykke Andersen, SDSN Bolivia
Boris Branisa C., ePC – Universidad Católica Boliviana
Fabián Calderón C., UPB

Abstract pending.

The Effect of Blackouts on Household Electrification Status: Evidence from Kenya

Raul Bajo Buenestado, University of Navarra

A number of countries in Sub Saharan Africa have recently deployed billions of dollars to improve their electricity infrastructure. However, aggregate data shows that the relative number of households with an electricity connection at home has barely increased. In this paper we study the role of blackouts to partially explain why there have been relatively few additional households with electricity access despite the increase in electrification expenditure. Using geo localized survey data from Kenya, we find that households that live in neighborhoods in which power outages are relatively more frequent are (at least) about 6–9 per cent less likely to have electricity at home. We also find that households that have electricity access but which experience frequent power outages are also less likely to purchase electrical appliances.

The impacts of household electrification in peri urban areas of Ecuador

Raul Jimenez Mori, IDB Invest

I investigate the effects of electricity service improvements on households
located in peri urban areas of Ecuador during the period 2014–2016. The identification strategy relies on double differences (DD) and instrumental variable (IV) approaches. As instrument, I use initial randomization of the project areas’ assignment to treatment and control. I present a series of tests and complementary results that support the causal interpretation of the estimations.
The evaluation includes a comprehensive causal chain encompassing quality of service, electricity consumption, time allocation of women and children, their safeness perception, as well as, household monetary income and adoption of energy efficient appliances, among others. The results indicate that the program has the expected effects on outcomes that are more directly related to the intervention, but they do not show robust impacts on more indirect outcomes.
The estimates support that the intervention improves the quality of electricity services as reported by the end users, increase electricity consumption, change time allocation within the household members, and increase adoption of energy efficiency. The effects on household income are significant under the DD strategy but they dilute after using IV, suggesting the presence of upward bias that would be expected from unobservable factors driving the allocation of the infrastructure improvements. Independently of the identification strategy, the analysis does not find effects on the perception of the quality of life, perception of safeness, frequency of respiratory illnesses, nor educational outcomes.