Contributions in Merlin

The maximum likelihood estimator for the regression coefficients, β, in a panel binary response model with fixed effects can be severely biased if N is large and T is small, a consequence of the incidental parameters problem. This has led to the development of conditional maximum likelihood estimators and, more recently, to estimators that remove the O(T–1) bias in β^. We add to this literature in two important ways. First, we focus on estimation of the fixed effects proper, as these have become increasingly important in applied work. Second, we build on a bias-reduction approach originally developed by Kosmidis and Firth (2009) for cross-section data, and show that in contrast to other proposals, the new estimator ensures finiteness of the fixed effects even in the absence of within-unit variation in the outcome. Results from a simulation study document favourable small sample properties. In an application to hospital data on patient readmission rates under the 2010 Affordable Care Act, we find that hospital fixed effects are strongly correlated across different treatment categories and on average higher for privately owned hospitals.

Audits are generally intended to monitor compliance with existing rules. However, audits can also create unintended impacts and incentives through the specific protocol by which they are executed. In particular, audits can discourage the use of complex administrative procedures with more rules for auditors to check. This paper investigates the effects of procurement audits on public entities' choice of purchase procedures in Chile. While the national procurement legislation tries to promote the use of more transparent and competitive auctions rather than discretionary direct contracts for selection of suppliers, auctions are significantly more complex and the audit protocol mechanically leads to more scrutiny and a higher probability of further investigation for auctions than for direct contracts. Using a regression discontinuity design based on a scoring rule of the National Comptroller Agency, we find that audits lead to a decrease in the use of auctions and a corresponding increase in the use of direct contracts. In order to further test the underlying mechanism, we develop a new approach to conduct subgroup analysis in regression discontinuity designs while holding other observables constant.

This paper proposes a post-model selection inference procedure, called targeted undersmoothing, designed to construct uniformly valid confidence sets for a broad class of functionals of sparse high-dimensional statistical models. These include dense functionals, which may potentially depend on all elements of an unknown high-dimensional parameter. The proposed confidence sets are based on an initially selected model and two additionally selected models, an upper model and a lower model, which enlarge the initially selected model. We illustrate application of the procedure in two empirical examples. The first example considers estimation of heterogeneous treatment effects using data from the Job Training Partnership Act of 1982, and the second example looks at estimating profitability from a mailing strategy based on estimated heterogeneous treatment effects in a direct mail marketing campaign. We also provide evidence on the finite sample performance of the proposed targeted undersmoothing procedure through a series of simulation experiments.

Sentiment indices, such as measures of consumer confidence, are often discussed as potential indicators of future investment, consumption and growth. However, documenting a causal relationship between consumer confidence and output — and understanding the precise nature of the relationship — using field data has been challenging. We rely on the high degree of control afforded by a laboratory setting to experimentally test a simple model of investment with complementarities and time-varying fundamentals. Our experiment manipulates the presence of aggregate confidence measures to test both how they reflect available information and how they influence future output. We find that an aggregate sentiment measure can be as effective as a highly precise exogenous public signal in coordinating behavior on more efficient equilibria. Furthermore, our analysis indicates that the confidence measure also impacts expectations by influencing beliefs about aggregate investment.

A growing theoretical and empirical literature shows that public recognition can lead employees to exert greater effort. However, status competition is also associated with excessive expenditure on status goods, greater likelihood of bankruptcy, and more risk taking by money managers. This paper examines the effects of recognition and status competition jointly. In particular, we focus on the spillover effects of public recognition on the performance and risk taking of peers. Using newly collected data on monthly “victory” scores of more than 5,000 German pilots during World War II, we find that status competition had important effects: After the German armed forces bulletin mentioned the accomplishments of a particular fighter pilot, his former peers performed considerably better. This outperformance varied across skill groups. When a former squadron peer was mentioned, the best pilots tried harder, scored more, and died no more frequently; in contrast, average pilots won only a few additional victories but died at a significantly higher rate. Hence our results show that the overall efficiency effect of nonfinancial rewards can be ambiguous in settings where both risk and output affect aggregate performance.

In a new survey we ask respondents, after a standard Subjective Well‐Being (SWB) question, if they can think of changes in their lives that would improve their SWB score. If the SWB score is just one argument among others in the respondents’ goals in life, they should easily find ways to improve it, at the expense of other dimensions they care about. Our results suggest that close to 90% of the respondents actually seek to maximize their SWB. The life satisfaction question appears the best contender as the “maximand” in the contest, before the ladder‐of‐life question and felt happiness. Among the other goals that people pursue and for which they are willing to sacrifice some of their SWB, the prominent appear to be about their relatives and about their future self.

The microfinance revolution has transformed access to financial services for low-income populations worldwide. As a result, it has become one of the most talked-about innovations in global development in recent decades. However, its expansion has not been without controversy. While many hailed it as a way to end world poverty and promote female empowerment, others condemned it as a disaster for the poor. Female empowerment has often been seen as one of the key promises of the industry. In part, this is based on the fact that more than 80% of its poorest clients, i.e., those who live on less than $1.25/day, are women. This paper discusses what we have learned so far about the potential and limits of microfinance and how insights from research and practice can help inform the industry's current products, policies and future developments.

Conditional heteroskedasticity can be exploited to identify the structural vector autoregressions (SVAR) but the implications for inference on structural impulse responses have not been investigated in detail yet. We consider the conditionally heteroskedastic SVAR-GARCH model and propose a bootstrap-based inference procedure on structural impulse responses. We compare the finite-sample properties of our bootstrap method with those of two competing bootstrap methods via extensive Monte Carlo simulations. We also present a three-step estimation procedure of the parameters of the SVAR-GARCH model that promises numerical stability even in scenarios with small sample sizes and/or large dimensions.

This paper investigates how high school gender composition affects students’ participation in STEM college studies. Using Danish administrative data, we exploit idiosyncratic within-school variation in gender composition. We find that having a larger proportion of female peers reduces women’s probability of enrolling in and graduating from STEM programs. Men’s STEM participation increases with more female peers present. In the long run, women exposed to more female peers earn less because they (1) are less likely to work in STEM occupations, and (2) have more children. Our findings show that the school peer environment has lasting effects on occupational sorting and the gender wage gap.

The paper reports the first experimental study on people’s fairness views on extreme income inequalities arising from winner-take-all reward structures. We find that the majority of participants consider extreme income inequality generated in winner-take-all situations as fair, independent of the winning margin. Spectators appear to endorse a “factual merit” fairness argument for no redistribution: the winner deserves all the earnings because these earnings were determined by his or her performance. Our findings shed light on the present political debate on redistribution, by suggesting that people may object less to certain types of extreme income inequality than commonly assumed.