« Back to Results
Poster Presentations of the Econometric Society
Sunday, Jan. 3, 2021
7:00 AM - 6:00 PM (EST)
Monday, Jan. 4, 2021
7:00 AM - 6:00 PM (EST)
Tuesday, Jan. 5, 2021
7:00 AM - 3:00 PM (EST)
University of Chicago
Sequential Learning under Informational Ambiguity
This paper studies a sequential learning problem where individuals are ambiguous about other people's signal structures. It finds that ambiguity has an important impact on social learning and provides new insights on the mechanism behind herding behavior. This paper claims that whether an information cascade occurs is a result of individuals' ambiguity level instead of specific statistical features of the actual signal processes as suggested by previous literature. When there is sufficient ambiguity, for all possible data-generating processes, an information cascade occurs almost surely. Moreover, a slight degree of ambiguity suffices to produce a cascade when signals are bounded and destroys full learning when signals are unbounded. As an extension, this paper also investigates the case where there is an outside option. It finds that an information cascade occurs on this outside option when there is sufficient ambiguity and individuals are ambiguity-averse.
A Quantitative Analysis of Distortions in Managerial Forecasts
This paper quantifies the economic costs of distortions in managerial forecasts. We match a unique managerial survey run by the Bank of Italy with administrative data on firm balance sheets and income statements. The resulting dataset allows us to observe a long panel of managerial forecast errors for a sample of firms representative of the Italian economy. We show that managerial forecast errors are positively and significantly autocorrelated. This persistence in forecast error is consistent with managerial underreaction to new information. To quantify the economic significance of this forecasting bias, we estimate a dynamic equilibrium model with heterogeneous firms and distorted expectations. The estimated model matches not only the persistence of forecast errors, but the empirical link between investment and managerial forecasts. Relative to a counterfactual with rational expectations, we find that managers exhibit large forecasting biases, which lead to significant distortions in firm-level investment. These distortions, however, imply limited loss in firm value. In general equilibrium, the estimated model leads to negligible aggregate efficiency losses from distorted forecasts.
The Cyclicality of Job Creation with Multiple Offers
In search of a job, a worker may receive more than one offer. This paper studies the interaction between multiple offers and job creation over the business cycle. Theoretically, I endogenize multiple offers and the wage offer distribution in a Diamond-Mortensen-Pissarides framework. The model predicts that both the fraction of new hires with multiple offers and the share of high-wage vacancies posted by firms are procyclical, the latter implies recessions reduce the quality of vacancies available to job seekers. Empirically, I present evidence consistent with both predictions. The findings contribute to the understanding of the sullying effect of recessions.
Multilateral Bargaining with Collective Proposal Control
I study a model of multilateral bargaining in which multiple proposers simultaneously make offers on several pies. I identify a novel source of inefficient delay unique to multilateral bargaining -- free-riding among proposers combined with the variability of proposal power. I establish that there exist stationary equilibria with delay and characterize the equilibrium agreement sets. In the worst equilibrium agents agree if and only if the proposal power is sufficiently concentrated. I compare the efficiency consequences of different voting rules, showing that voting rules requiring approval by greater majorities lead to more delay in the worst equilibrium. Agents' payoffs in the worst equilibrium are maximized if the distribution of the proposal power is such that power is concentrated to a moderate extent. Finally, I show that if the proposal power can be divided finely among the agents, then there are distributions of power such that in the worst equilibrium, as the number of agents grows large, delay consumes almost all surplus.
Preference for Simplicity
This paper introduces and axiomatizes representations in which the agent assesses a lottery less favorably if it contains more outcomes. These representations, which we term simplicity theory, are motivated by experimental evidence. They capture as special cases expected utility, the certainty effect, and a range of laboratory and em- pirical phenomena. We compare simplicity theory to existing theories including expected utility theory and prospect theory. We provide parametric examples of simplicity theory, and relate the theory to applications including risk aversion studies, cognitive psychology research, and household financial behavior.
The Rise of a Network: Spillover of Political Patronage and Cronyism to the Private Sector
We document that networks that gain access to political power and use it for patronage appointments also gain control over resource allocation in the private sector. Specifically, following a presidential election in Korea, the president appoints members of his network into important positions in government, and private banks respond by appointing executives from the same network to establish links to the administration. Consequently, firms linked to the network obtain more credit at a lower rate from government and private banks alike, despite higher default rates. Micro-level data on loans and variation in network links for the same firm across lenders over time sharpen the interpretation of our results. In a parsimonious model, we show that efficiency costs are higher when government and private banks are controlled by the same group rather than different groups: in-group firms invest in more unprofitable projects, whereas out-group firms lack funding for highly profitable investments.
Robust Opinion Aggregation and Its Dynamics
We study agents in a social network who receive initial noisy signals about a fundamental parameter and then, in each period, solve a robust non-parametric estimation problem given their previous information and the most recent estimates of their neighbors. The resulting robust opinion aggregators are characterized by simple functional properties: normalization, monotonicity, and translation invariance. These aggregators admit the linear DeGroot's model as a particular parametric specification. However, robust opinion aggregators allow for additional features such as overweighting/underweighting of extreme opinions, confirmatory bias, as well as discarding information obtained from sources perceived as redundant. We show that under this general model, it is still possible to link the long-run behavior of the opinions to the structure of the underlying network. In particular, we provide sufficient conditions for convergence and consensus and we offer some bounds on the rate of convergence. In some parametric cases, we derive the influence of the agents on the limit opinions and we stress how it depends on their centrality as well as on their initial signals. Finally, we study sufficient conditions under which a large society learns the true parameter while also highlighting why this property may fail.
Super PACs and Political Polarization: If You Can’t Convince Them, Displace Them
We investigate the role of the introduction of Super PACs on political polarization, by exploiting a natural experiment provided by the 2010 Citizens-United supreme court ruling. This decision eliminated the limitations on independent campaign advertising, and de-facto lead to a sizable increase in the amount spent on negative ads. We first show that the ruling is correlated with political polarization establishing that there is a time break that can be explained, and then find that this discontinuity is associated with a significant change in the number of negative ads. Finally, we provide evidence that independent expenditures against candidates completely explain the break. These results are consistent with a theoretical framework that relies on the intuition that negative advertisements tend to have a composition effect on electoral turnout. Indeed, campaigns that focus on the negative attributes of competing politicians tend to drive away moderate voters, while at the same time mobilizing partisan voters that are more prone to follow their party allegiance.
Civil War Battles and Credit Markets
I study how a shock in culture and trust in the past still affects mortgage lending today. More specifically, I investigate the long run effects of the American Civil War, a defining event for culture in American history, on current mortgage lending approval between 2000 and 2011. Using a spatial regression discontinuity design, exploiting the random occurrences of battles during the Civil War, I find that location matters for credit extension: Being located in a county in which a battle took place fosters the probability of loan approval. However, applicants from minority groups, such as African Americans, have a significant lower probability to obtain a mortgage loan in counties where a battlefield during the Civil War was located compared to similar applicants in adjacent non-battle counties. Conditional upon approval they also receive lower loan amounts. I show that a channel through which this battle effect persists is culture: Counties in which soldiers actively fought during the Civil War show higher levels of social capital today. Additionally, I find that remembrances of Civil War battles are important for the persistency of local social capital: Those battle counties that actively remember through re-enactment groups have even higher social capital today compared to those battle counties that do not. Moreover, applicants from minority groups have a significant lower probability to see their application currently being approved in these counties compared to minority applicants from battle counties that do not actively remember through such groups. Furthermore, conditional upon approval, they are also granted a lower loan amount, suggesting a possible ‘dark side’ of social capital. These effects are reinforced in counties where the Union won a battle but mitigated by minority bank presence.
Energy Political Uncertainty, Investment Opportunities, and Stock Returns
In this paper I study the relation among political uncertainty regarding future energy policies, corporate investment decisions, and stock returns. Energy political uncertainty (EnPU) coincides with investment-specific technological shocks, capturing aggregate investment and lower future marginal utility. Periods of high energy political uncertainty negatively capture expected return variation not explained by changes in risk aversion. In anticipation of lower expected returns, investors demand growth companies whose cash-flows covary positively with innovations on uncertainty relative to value companies. These results are robust to the way in which investors form expectations about political decisions, as well as to industry specific dynamics. My results suggest that uncertainty about energy policies increases the premium of the option to invest of growth companies.
The Real Effects of Environmental Activist Investing
We study the real effects of environmental activist investing. Using plant-level data, we find that targeted firms reduce their toxic releases, greenhouse gas emissions, and cancer-causing pollution. Improvements in air quality within a one-mile radius of targeted plants suggest potentially important externalities to local economies. These improvements come through increased capital expenditures on new abatement initiatives. We rule out alternative explanations of decline in production, reporting biases, and forms of selection, while also providing evidence supporting the external validity of environmental activism. Overall, our study suggests that engagements are an effective tool for long-term shareholders to address climate change risks.
Market Discipline under Financial Contagion
We investigate the disciplining eﬀect of debt on bank management under a situation where ﬁnancial contagion occurs. Theory analyzing market discipline, which deals with a single bank, suggests that bankers become "more creditor friendly" by holding demand deposit or short-term deb (e.g. Calomiris and Kahn 1991; Diamond and Rajan 2012). However, when there is also a ﬁnancial market other than a bank and when these are interconnected through a contagion of ﬁnancial crises, an alternative eﬀect arises: namely, a banker’s incentive to make an eﬀort is aﬀected by the behavior of speculators in the ﬁnancial market. We show that, when the speculators have more precise information about fundamentals of the economy than that of creditors in the bank, demand deposit or short-term debt induces the bankers to "act against creditors’ interests" under some conditions.
Evidence Acquisition and Voluntary Disclosure
A sender seeks hard evidence to persuade a receiver to take a certain action. There is uncertainty about whether the sender obtains evidence. If she does, she can choose to disclose it or pretend to not have obtained it. When the probability of obtaining information is low, we show that the optimal evidence structure is a binary certification: all it reveals is whether the (continuous) state of the world is above or below a certain threshold. Moreover, the set of low states that are concealed is non-monotone in the probability of obtaining evidence. When binary structures are optimal, higher uncertainty leads to less pooling at the bottom because the sender uses binary certification to commit to disclose evidence more often.
Impact of British Colonial Gender Legal Reforms: Evidence from Child Marriage Abolition Act, 1929
The British colonial government raised the minimum age at the first marriage of females to 14 years in British India in 1929. The law had two distinct features: it was announced in 1929 but implemented six months later in 1930, and the law applied to only British India, which was directly ruled by the British government but not to Princely States (Indian Native States), which were under their indirect control. Using the Princely States in colonial India as a control group, we employ a difference-in-differences strategy to estimate the causal impact of the abolition of child marriage on underage female marriage in regions affected by the law. Analyzing historical census data from 1911-1931, we find an immediate unintended anticipation effect of the law that increased female child marriages in the affected regions in 1931, followed by a sharp decline of female child marriages in the affected region post-independence in 1961-1981. We further use three independent nationally representative data sets on female education and marriage to show a long-term decline in child marriages and an increase in educational attainment among women in affected regions. The short- vs. long-run contrast provides evidence of the efficacy of the legal reform. Although the native population in the short run took actions to preempt the law, the long-term results reveal that the colonial government succeeded in changing practice and possibly norms. Our findings speak to the importance of a very long follow-up in the study of interventions that promote social change.
Sharing Risk to Avoid Tragedy: Informal Insurance and Irrigation in Village Economies
I present a model of interaction between risk sharing and co-operation over irrigational investments in presence of limited commitment constraints and apply it to the case of farmers in rural India. I demonstrate that if access to irrigation can be regulated by villagers, the two institutions reinforce each other. However, if benefits of such investments are non-excludable (as is the case with provision by central authorities), they may harm local co-operation. Using the ICRISAT panel, I provide empirical support for the mechanism. Quantitative evaluation demonstrates significant reinforcement between the two institutions and economic losses due to sub-optimal irrigation management.
Robots and Reshoring: Evidence from Mexican Labor Markets
Robots in advanced economies have the potential to reduce employment in offshoring countries by fueling reshoring. Using robots instead of humans for production may lower the relative cost of domestic production and, in turn, reduce demand for imports from offshoring countries. I analyze the impact of robots on employment in an offshoring country, using data from Mexican local labor markets between 1990 and 2015. A recent literature estimates the effect of robots on local employment by regressing the change in employment on exposure to domestic robots in local labor markets. I construct a similar measure of exposure to foreign robots, based on the initial geographic distribution of exports producing employment across industries, industry-level robot adoption in the US, and a US industry's initial reliance on Mexican imports. To purge results from endogeneity, I use robot adoption in the rest of the world and an index of offshoring as instruments for robot adoption in the US and the share of Mexican imports, respectively. Using these instruments, I show that US robots have a sizeable, negative impact on employment in Mexico. This negative effect is stronger for men than for women, and strongest for low-educated machine operators in the manufacturing sector. Consistent with reshoring as a mechanism, I find that the employment effect is mirrored in similarly large reductions in Mexican exports and exports producing plants.
Spillovers and Redistribution through Intra-firm Networks: The Product Replacement Channel
We study how regional shocks spill over across US local markets through intra-firm market networks and explore how such spillovers reshape household welfare across regions. We identify spillovers by linking data on barcode-region-level prices and quantities with producer-level information and by exploiting variation in firms’ exposure to sudden differential drops in local house prices. We find that a firm’s local sales decrease in response to a direct negative local demand shock and do so more strongly to indirect negative demand shocks originating in its other markets. The intra-firm cross-market spillover effects occur because (i) firms replace higher-value products—higher sales per product, unit price, and organic sales share—with lower-value products in response to negative demand shocks, and (ii) such product replacements are synchronized across markets within each firm. Counterfactual analysis using multiregion model with endogenous quality adjustment shows that our channel generates a novel and economically sizable regional redistribution effect during the Great Recession.
Valuing Domestic Transport Infrastructure: A View from the Route Choice of Exporters
A key input to quantitative evaluations of transport infrastructure projects is their impact on transport costs. This paper proposes a new method of estimating this impact relying on the widely accessible customs data: by using the route choice of exporters. We combine our method with a spatial equilibrium model to study the aggregate effects of the massive expressway construction in China between 1999 and 2010. We find that the construction brings 5.6% welfare gains, implying a net return to investment of 180%. Our analysis also produces some intermediate output of independent interest, for example, a time-varying IV for city-sector export.
Managers and Productivity in the Public Sector
This paper studies the productivity impacts of managers in the public sector using novel administrative data containing an output-based measure of productivity of public offices. Exploiting the rotation of managers across sites, I find that a one standard deviation better manager increases office productivity by 10%. These gains are driven primarily by the exit of older workers who retire when more productive managers take over. Empowering managers to directly change payrolls may generate large benefits to efficiency. Absent such civil service reforms, I use these estimates to evaluate the optimal allocation of managers to offices. I find that assigning better managers to the largest and most productive offices would increase output by at least 6.9%.
Technical Change and Superstar Effects: Evidence from the Roll-Out of Television
Technical change that improves economies of scale can benefit the most talented individuals at the expense of everyone else. I test this classic Rosen "superstar model" in the market for entertainers where the historic roll-out of television led to a natural experiment in scale-related technological change. When a local TV station launched, top entertainers' audiences increased 4-fold, leading to a 50% increase in the top-percentile income share, while reducing employment and incomes of lower-level talents. The results confirm the predictions of "superstar effects" and are inconsistent with canonical models of skill-biased technological change.
The Signaling Role of Parental Leave
This paper provides empirical tests of signaling theory in the context of workers choosing to forgo paid parental leave to signal value to employers. Using administrative data from Denmark and an unanticipated increase in the maximum allowed duration of parental leave, I show how signaling contributes to a divergence in wages due to the information that workers’ choices convey upon a leave extension. In contrast to human capital theory, an individual can take longer leave but gain in wages as long as their relative leave position in the population decreases. The results demonstrate unanticipated impacts of parental leave policies and highlight the importance of asymmetric information in the labor market.
Equilibrium Effects of Pay Transparency in a Simple Labor Market
The public discourse around pay transparency has focused on the direct effect: how workers seek to rectify newly disclosed pay inequities through renegotiations. The question of how wage-setting, bargaining, and hiring practices change with higher transparency has received less attention. To study these outcomes, we combine a dynamic wage-bargaining model with data from online labor markets for low-skill, temporary jobs. We exploit naturally occurring variation in pay transparency as well as experimentally-induced variation. Wages are more equal, but lower under transparency. An increase in transparency raises the hiring rate when workers have sufficient bargaining power. Employer profits rise with transparency, increasing 27% in a field experiment. A key insight is that increasing transparency decreases the bargaining power of workers, as employers credibly refuse to pay high wages to any one worker to avoid costly renegotiations with others. We discuss implications for the gender wage gap and employers' endogenous transparency choices.
Firm Productivity and Cities: The Case of Colombia
We study the determinants of firm-level productivity in Colombia. We are interested in the effects of agglomeration forces that explain why manufacturing economic activity tends to concentrate geographically as well as the effect of forces that can hurt productivity in high-density areas. An unfortunate major example of the latter forces in Colombia is its high levels of crime and terrorist attacks. We carry out this study by exploiting two very rich data sources: a firm-level panel that allows us to estimate firm-level productivity and a panel of municipality characteristics. To address selection and endogeneity issues in the estimation of firm-level productivity we use a control function approach. Our main findings are the following. First, scale economies do not seem to affect firm-level productivity. Second, we do find evidence of location economies (industrial specialization has a positive effect on productivity). Industrial variety, on the other hand, lowers productivity. It seems that firms in Colombia benefit from forming clusters and locating in cities with less industrial variety. We also find non-trivial effects of city fiscal performance, education level and quality, and crime and violent attacks.
Steering via Algorithmic Recommendations
This paper studies whether market structure affects algorithmic recommendations in dominant platforms. We focus on the dual role of Amazon.com---as a platform owner and retailer. We find that products sold by Amazon receive substantially more ``Frequently Bought Together” recommendations across product categories and popularity deciles. To establish causality, we exploit within-product variation generated by Amazon stockouts. We find that when Amazon is out of stock, the identical product sold by third-party sellers receives 8% fewer recommendations. The pattern can be explained by economic incentives of steering and cannot be explained by consumer preference. Furthermore, the steering lowers recommendation efficiency.
The Information Driven Financial Accelerator
Imperfect information in credit markets is a quantitatively important source of macroeconomic fragility. We calibrate a dynamic model with uninformed debt investors. A deterioration in the profit outlook makes investors pessimistic about firm creditworthiness. In turn, firms perceive that debt is underpriced and cut back investment. We show that: 1) the model matches the size and cyclical variation of credit spreads; 2) imperfect information accounts for about half of the spike in spreads and one-fifth of the contraction in aggregate investment during the US financial crisis; 3) the economic costs of imperfect information for firm value and investment are substantial.
Inequality, Taxation, and Sovereign Default Risk
Government debt repayment relies on its tax revenue, the amount of which crucially depends on the micro features of the economy. This paper argues that income inequality and its interactions with migration significantly affect sovereign default risk. With high income inequality, the government imposes progressive taxes, which redistribute income but discourage labor supply and induce emigration, eroding the tax base and government's ability to repay debt. I develop a sovereign default model with endogenous non-linear taxation and heterogeneous labor with labor mobility to match those patterns. In the model, government chooses the optimal combination of tax and debt, considering its impact on the households' decisions and welfare. With the estimated model, I find that income inequality and its interactions with migration explain 23% of the variations in spreads across the states in the US.
Screening with Price and Expertise
We consider competitive asset markets where quality is heterogeneous and buyers can use their expertise to evaluate quality. After paying an entry cost, each buyer can set up a potential market by posting a fixed price and recommending certain quality distribution and market tightness. Each seller can search in any potential market. When a match is formed, the buyer obtains a noisy quality signal about her counterparty and decides whether to accept his asset. The equilibrium requires all potential markets to deliver market utility to sellers in the support of the recommended quality distribution and, for active markets, quality distribution and market tightness are consistent with strategies. We find that the equilibrium of the model is characterized by a house race between screening with price and screening with expertise. When buyers' signal is inaccurate or when their entry cost is high, screening with price dominates, and the equilibrium is comprised of separating markets only. However, if the buyers' signal is accurate or their entry cost is low, a pooling market will arise where buyers screen with expertise by rejecting an asset if its quality signal is below the bar. We also analyze the scenario where buyers differ in their levels of expertise. In this case, buyers with high expertise will specialize in the pooling market, while buyers with low expertise may set up a separating market or a pooling market at a higher price.
Redistribution, Sovereign Debt, and Optimal Taxation
This paper proposes a theory of external debt sustainability based on the government’s motive for redistribution. I study a small open economy model in which taxes are distortionary and the government has a redistributive concern and faces endogenous borrowing constraints due to its lack of commitment. Given these borrowing constraints, the value of financial autarky determines the sustainable level of debt. Financial autarky is endogenously costly because, in this case, redistribution requires high labor taxes, which distort labor supply and reduce the economy's efficiency. Having access to external financing allows the government to have more redistribution, measured as the differences in individual utilities, than in financial autarky at the same level of efficiency cost. Quantitatively, the theory can account for the external debt’s recent buildup in Italy and is consistent with the positive correlation between pre-tax income inequality and external debt across countries and time periods. In response to a negative productivity shock, the optimal austerity policies are increasing external borrowing and redistribution while reducing redistribution to repay debt in the future. The magnitude of these responses varies with the underlying wage inequality.
Heterogeneous Passthrough from TFP to Wages
What is the impact of firms’ productivity shocks on workers’ labor earnings? To answer this question, we propose a new method to identify firms’ productivity shocks that combines a nonparametric production function estimation method with a set of two-way fixed effect regressions to control for differences in labor quality across firms. We apply this method to an employer-employee matched panel dataset that encompasses the entire population of workers and firms in Denmark between 1995 and 2010. Our dataset allows us to separately study workers that stay in the firm across consecutive periods from those that transition between firms, control for workers’ endogenous job mobility decisions, and to investigate how the passthrough from firms’ shocks to wages varies across narrow population groups. We find an elasticity of workers’ hourly wages to firms’ productivity of 0.08. This implies that a positive shock to firms' productivity of one standard deviation generates an increase of $1,100 US dollars in annual wages for the average worker in Denmark. We also find that both persistent and transitory shocks to firms are passed on to wages and that there is a marked asymmetry in passthrough from positive and negative productivity shocks. In fact, after controlling for selection, the elasticity of hourly wages to a negative productivity shock is twice that of a positive productivity shock of the same magnitude. This suggests that workers are more exposed to negative than to positive shocks to firms. Furthermore, we find that the changes in wages due to variation in firm productivity are quite persistent and do not dissipate even five years after the shock. By looking at the heterogeneity of passthrough across firms and workers groups we provide insights about the theoretical mechanisms that could explain the patterns of passthrough we observe in the data.
The Perils of Tracking r-Star
The natural rate of interest (r-star) has been a critical benchmark for monetary policymaking recently. We show that there exist fiscal limits to a monetary policy rule that targets r-star. When monetary policy is constrained by fiscal sustainability concerns, an interest rate rule that tracks r-star generates large and persistent movements in inflation and output gap, thereby producing macroeconomic instability and welfare losses. The mechanism operates through the government budget constraint, as interest rate changes affect the value of government debt and, thereby, inflation. This leads to perils of tracking r-star, in a model widely used for monetary policy analysis.
Identification and Estimation of Possibly Non-Invertible SVARMA Models: A New Parametrization
This paper deals with parameterisation, identifiability, and maximum likelihood (ML) estimation of possibly non-invertible structural vector autoregressive moving average (SVARMA) models driven by independent and non-Gaussian shocks. We introduce a new parameterisation of the MA polynomial matrix based on the Wiener-Hopf factorisation (WHF) and show that the model is identified in this parametrisation for a generic set in the parameter space (when certain just-identifying restrictions are imposed).When the SVARMA model is driven by Gaussian errors, neither the static shock transmission matrix, nor the location of the determinantal zeros of the MA polynomial matrix can be identified without imposing further identifying restrictions on the parameters.We characterise the classes of observational equivalence with respect to second moment information at different stages of the modelling process. Subsequently, cross-sectional and temporal independence and non-Gaussianity of the shocks is used to solve these identifiability problems and identify the true root location of the MA polynomial matrix as well as the static shock transmission matrix (up to permutation and scaling). Typically imposed identifying restrictions on the shock transmission matrix as well as on the determinantal root location are made testable. Furthermore, we provide low level conditions for asymptotic normality of the ML estimator and analytic expressions for the score and the information matrix. As application, we investigate the impact of a government spending shock on consumption and real wages without including constructed news shock variables like in Ramey's analyses. This and further analyses are implemented in a well documented R-package.
The Effects of Working While in School: Evidence from Uruguayan Lotteries
Shall we encourage students to work while in school? We provide evidence by leveraging a one-year work-study program that randomizes job offers among students in Uruguay. Using social security data matched to over 120,000 applicants, we estimate an increase of 9% in earnings and of 2 percentage points in enrollment over the four post-program years for treated youth. Survey data indicate that enrolled participants reduce study time, but this does not translate into lower grades. Students mainly substitute leisure and household chores with work. The earnings effect is related to the work experience and the transferability of skills acquired in program jobs.
Labour Supply and the Pension Contribution-Benefit Link
We estimate the labour supply response to a reform that switched the Polish pension system from a defined benefit (DB) to a notional defined contribution (NDC) scheme. The key aspect of this reform was to change the link between current pension contributions and future pension benefits. Only those born after December 31st 1948 were affected, creating a sharp cohort-based discontinuity in the contribution-benefit link for individuals in their 50s. Using administrative data on the universe of Polish taxpayers, we examine labor supply responses to the reform at ages 51-54, around 11-15 years before the normal retirement age. In line with the change in incentives, we find that the employment rate decreased by 2% in response to the policy change, which is consistent with an extensive margin elasticity of 0.17. These responses are driven by regions where the incentives had changed the most. Our results imply that changing the contribution-benefit link of public pensions can alleviate the labor supply distortions caused by security contributions.
Matching Points: Supplementing Instruments with Covariates in Triangular Models
We consider triangular models with a discrete endogenous variable and an instrumental variable (IV) taking on fewer values. Addressing the failure of the order condition, we develop the first approach to restore identification for both separable and nonseparable models in this case by supplementing the IV with covariates, allowed to enter the model in an arbitrary way. For the separable model, we show that it satisfies a system of linear equations, yielding a simple identification condition and a closed- form estimator. For the nonseparable model, we develop a new identification argument by exploiting its continuity and monotonicity, leading to weak sufficient conditions for global identification. Built on it, we propose a uniformly consistent and asymptotically normal sieve estimator. We apply our approach to an empirical application of the return to education with a binary IV. Though under-identified by the IV alone, we obtain results consistent with the literature using our approach. We also illustrate the applicability of our approach via an application of preschool program selection where the supplementation procedure fails.
Nonparametric Bounds on Treatment Effects with Imperfect Instruments
This paper extends the identification results in Nevo and Rosen (2012) to nonparametric models. We derive nonparametric sharp bounds on the average treatment effect when an imperfect instrument is available. As in Nevo and Rosen (2012), we assume that the sign of correlation between the imperfect instrument and the unobserved latent variables is the same as the correlation between the endogenous variable and the latent variables. We introduce the concept of comonotone instrumental variable, which satisfies this assumption. We show how the assumption that the imperfect instrument is less endogenous than the treatment variable can help tighten the bounds. We also use the monotone treatment response assumption to get tighter bounds. The identified set can be written in the form of intersection bounds, which is more conducive to inference. We illustrate our methodology using the National Longitudinal Survey of Young Men data to estimate returns to schooling.
Peer Effects in Random Consideration Sets
This paper develops a dynamic model of discrete choice that incorporates peer effects into random consideration sets. We characterize the equilibrium behavior and study the empirical content of the model. In our setup, changes in the choices of friends induce changes in the distribution of the consideration sets. We exploit this variation to recover the ranking of preferences, the attention mechanisms, and the network connections. These identification results allow unrestricted heterogeneity across people and do not rely on variation of either covariates or the set of available options. Our methodology leads to a maximum-likelihood estimator that performs well in simulations.
Identification of Auction Models Using Order Statistics
Auction data often fail to record all bids or all relevant factors that shift bidder values, which complicates the identification of the underlying value distribution. This paper considers the identification of independent private value auctions with nonseparable finite unobserved heterogeneity (UH) using multiple order statistics of bids. Usual measurement error approaches are inapplicable due to dependence among order statistics. We provide a set of positive results. First, we show that auctions with symmetric bidders are identifiable using three consecutive order statistics or two consecutive ones with an instrument. Second, we allow for two dimensions unobserved heterogeneity including the level of competition unknown. Third, we then extend our identification results for auctions with asymmetric bidders. Lastly, we apply our methods to U.S. Forest Service timber auctions and find that ignoring UH reduces both bidder and auctioneer surplus.
The Flexible Inverse Logit Model
This paper develops the flexible inverse logit model, a structural inverse demand model which is able to describe the behavior of heterogeneous, utility-maximizing consumers choosing from a choice set of possibly many products that are differentiated in a way that is both observed and unobserved by the modeller. The FIL model is easy to estimate by linear instrumental variables regression to deal with the endogeneity issues of prices and market shates due to the modelling of unobserved product differentiation. Furthermore, the FIL model accommodates rich substitution patterns, including complementarity in demand. In particular, simulations show that it is able to match the substitution patterns of the random coefficient logit pretty well and to get quite right predictions of a merger's price and share effects.
Estimating Spillover Effects with Matched Data
Social interactions often play a key role in determining the impact of policies, but measuring the magnitude of spillover effects empirically is notoriously challenging because, in most applications, a person's relationships are likely to reflect her own characteristics (homophily), and people who are connected are likely to be affected by the same shocks (common factors). In addition, a significant share of social interactions is likely to be at play through variables that are not observed by the researcher. When matched data are used, observations corresponding to the same cross-sectional units (e.g., firms or students) can be linked over time, and a cross-sectional unit's relationships (e.g., co-workers or classmates) are indexed in each time period. We show that comparisons along the mobility network of a matched dataset (e.g., workers employed by different firms over time) can be used to measure the importance of social interactions in the presence of flexible patterns of selection on unobservables and common factors, and even if social interactions only occur through unobservables. We apply our results to estimate the importance of peer effects in determining the propensity of elementary school students with exceptionalities to commit disciplinary infractions.
Identification of Random Coefficient Latent Utility Models
This paper provides nonparametric identification results for random coefficient distributions in perturbed utility models. We cover discrete choice and models of multiple purchases. We establish identification using variation in mean quantities. The results apply even when an analyst observes only aggregate demands but not whether goods are chosen together. The identification results use exclusion restrictions on covariates and independence between random slope coefficients and random intercepts. We do not require regressors to have large supports or parametric assumptions.
The Converse Envelope Theorem
I prove an envelope theorem with a converse: the envelope formula is equivalent to a first-order condition. Like Milgrom and Segal’s (2002) envelope theorem, my result requires no structure on the choice set. I use the converse envelope theorem to extend to abstract outcomes the canonical result in mechanism design that any increasing allocation is implementable, and apply this to selling information.