When less conditioning provides better estimates: overcontrol and endogenous selection biases in research on intergenerational mobility

The counterfactual approach to causality has become the dominant approach to understand causality in contemporary social science research. Whilst most sociologists are aware that unobserved, confounding variables may bias the estimates of causal effects (omitted variable bias), the threats of overcontrol and endogenous selection biases are less well known. In particular, widely used practices in research on intergenerational mobility are affected by these biases. I review four of these practices from the viewpoint of the counterfactual approach to causality and show why overcontrol and endogenous selection biases arise when these practices are implemented. I use data from the German Socio-Economic Panel Study (SOEP) to demonstrate the practical consequences of these biases for conclusions about intergenerational mobility. I conclude that future research on intergenerational mobility should reflect more upon the possibilities of bias introduced by conditioning on variables.

The deterrence effect of real-world operational tax audits on self-employed taxpayers: evidence from Italy

We use a large administrative tax-returns panel dataset merged with a tax audit database to estimate the effect of real-world operational tax audits on subsequent tax behavior of a large sample of Italian self-employed taxpayers. Results from operational audits do not suffer from the fact that taxpayers are aware that they have been randomly selected for research purposes and, then, such audits are viewed as more of a signal about true audit rates by the taxpayer. Our empirical approach relies on fixed-effects difference-in-difference comparisons with an ex-ante matched sample of non-audited taxpayers. To address concerns about the endogenous selection into audit, we provide evidence for the common trends assumption. We find a positive and lasting effect of audits on subsequent reported income. However, in line with theoretical predictions, taxpayers do not increase tax compliance when the tax authority does not assess a positive additional income. Our results are robust to a variety of specifications and samples.

Give it Another Try: What are the Effects of a Job Creation Scheme Especially Designed for Hard-to-Place Workers?

Previous evaluations of job creation schemes (JCS) reveal mostly negative employment effects, mainly due to inherent lock-in effects. In this paper, we assess the impact of an innovative JCS that employs a pre-selection mechanism to target programme participation on unemployed job seekers with very low integration chances, hereby reducing possible lock-in effects. Relying on high-quality administrative as well as survey data, we conduct regression-adjusted matching analyses to estimate the programme effect on integration into regular employment. Our results show that the programme did not succeed to foster labour market integration, but still entails remarkably negative employment effects in the first years after participation. We argue that this results from a principal-agent problem at the last step of the selection mechanism that may have led to cream-skimming rather than targeting on very hard to place workers. However, supplementary analyses reveal that negative effects can be avoided for subgroups with very poor employment chances in case of non-participation. These results are robust to the use of different matching estimators and definitions of non-participation. The inclusion of usually unobservable survey variables as well as placebo tests based on past employment outcomes refute concerns about endogenous selection. From a policy-perspective, these findings imply that targeting JCS on workers with very low integration chances is a key factor to avoid negative employment effects found in previous evaluations. At the methodological level, our analyses add to recent literature that assesses the credibility of non-experimental evaluations based on high-quality administrative data.

Micro-Level Misallocation and Selection

How large are the aggregate productivity losses from the misallocation of resources across firms? With endogenous selection, microfrictions can induce extensive margin misallocation among firms: too many unproductive firms are active (Zombies), and too many productive firms are inactive (Shadows). Therefore, the same set of measured distortions potentially induces much larger aggregate productivity losses, as the composition of firms is shifted toward unproductive active firms. I develop and calibrate a model with plant-level microdata for Indonesia to quantify aggregate welfare in the presence of extensive margin misallocation. My estimates show that selection can magnify aggregate TFP losses from microdistortions by over 40 percent compared to existing estimates. Realistic values of measurement error even increase the relative importance of extensive margin misallocation. (JEL D22, D24, E23, J24, J31, O14, O15)

How to Sell a Data Set? Pricing Policies for Data Monetization

The wide variety of pricing policies used in practice by data sellers suggests that there are significant challenges in pricing data sets. In this paper, we develop a utility framework that is appropriate for data buyers and the corresponding pricing of the data by the data seller. Buyers interested in purchasing a data set have private valuations in two aspects—their ideal record that they value the most, and the rate at which their valuation for the records in the data set decays as they differ from the buyers’ ideal record. The seller allows individual buyers to filter the data set and select the records that are of interest to them. The multidimensional private information of the buyers coupled with the endogenous selection of records makes the seller’s problem of optimally pricing the data set a challenging one. We formulate a tractable model and successfully exploit its special structure to obtain optimal and near-optimal data-selling mechanisms. Specifically, we provide insights into the conditions under which a commonly used mechanism—namely, a price-quantity schedule—is optimal for the data seller. When the conditions leading to the optimality of a price-quantity schedule do not hold, we show that the optimal price-quantity schedule offers an attractive worst-case guarantee relative to an optimal mechanism. Further, we numerically solve for the optimal mechanism and show that the actual performance of two simple and well-known price-quantity schedules—namely, two-part tariff and two-block tariff—is near optimal. We also quantify the value to the seller from allowing buyers to filter the data set.

Does the Opening of High-Speed Railway Lines Reduce the Carbon Intensity of China’s Resource-Based Cities?

The carbon intensity of China’s resource-based cities (RBCs) is much higher than the national average due to their relatively intensive mode of development. Low carbon transformation of RBCs is an important way to achieve the goal of reaching the carbon emissions peak in 2030. Based on the panel data from 116 RBCs in China from 2003 to 2018, this study takes the opening of high-speed railway (HSR) lines as a quasi-experiment, using a time-varying difference-in-difference (DID) model to empirically evaluate the impact of an HSR line on reducing the carbon intensity of RBCs. The results show that the opening of an HSR line can reduce the carbon intensity of RBCs, and this was still true after considering the possibility of problems with endogenous selection bias and after applying the relevant robustness tests. The opening of an HSR line is found to have a significant reducing effect on the carbon intensity of different types of RBC, and the decline in the carbon intensity of coal-based cities is found to be the greatest. Promoting migration of RBCs with HSR lines is found to be an effective intermediary way of reducing their carbon intensity.

Adverse and Advantageous Selection in the Laboratory

We study two-player games where one-sided asymmetric information can lead to either adverse or advantageous selection. We contrast behavior in these games with settings where both players are uninformed. We find stark differences, suggesting that subjects do account for endogenous selection effects. Removing strategic uncertainty increases the fraction of subjects who account for selection. Subjects respond more to adverse than advantageous selection. Using additional treatments where we vary payoff feedback, we connect this difference to learning. We also observe a significant fraction of subjects who appear to understand selection effects but do not apply that knowledge. (JEL C92, D82, D91)

Gene-environment dependencies lead to collider bias in models with polygenic scores

The application of polygenic scores has transformed our ability to investigate whether and how genetic and environmental factors jointly contribute to the variation of complex traits. Modelling the complex interplay between genes and environment, however, raises serious methodological challenges. Here we illustrate the largely unrecognised impact of gene-environment dependencies on the identification of the effects of genes and their variation across environments. We show that controlling for heritable covariates in regression models that include polygenic scores as independent variables introduces endogenous selection bias when one or more of these covariates depends on unmeasured factors that also affect the outcome. This results in the problem of conditioning on a collider, which in turn leads to spurious associations and effect sizes. Using graphical and simulation methods we demonstrate that the degree of bias depends on the strength of the gene-covariate correlation and of hidden heterogeneity linking covariates with outcomes, regardless of whether the main analytic focus is mediation, confounding, or gene × covariate (commonly gene × environment) interactions. We offer potential solutions, highlighting the importance of causal inference. We also urge further caution when fitting and interpreting models with polygenic scores and non-exogenous environments or phenotypes and demonstrate how spurious associations are likely to arise, advancing our understanding of such results.

The Impact of Initial Public Offerings on Firms’ Performance: Disentangling Treatment from Self-Selection Effects

Using a unique sample of privately held and firms that went public on the European and Asian Stock Exchanges between 2007 and 2011, we investigate the IPO’s impact on the firms’ performance after correcting for endogenous selection and by disentangling equity issues effects from other effects. We find that companies that are going public are more profitable than their matched private firms, while they experience a decrease in profitability over the post-IPO period. These results are resilient to different empirical strategies that address selection bias. Second, after disentangling equity issues effects from other effects, we observe a continuous decline in firms’ profitability in each individual year following the IPO year. JEL classification numbers: G10, G30, G32, L25. Keywords: IPOs, Private firms, Profitability, Selection bias.