Contributions in Merlin

Transitivity is perhaps the most fundamental choice axiom and, therefore, almost all economic models assume that preferences are transitive. The empirical literature has regularly documented violations of transitivity, but these violations pose little problem as long as they are simply a result of somewhat-noisy decision making and not a reflection of the deterministic part of individuals’ preferences. However, what if transitivity violations reflect individuals’ genuinely nontransitive preferences? And how can we separate nontransitive preferences from noise-generated transitivity violations – a problem that so far appears unresolved? Here we tackle these fundamental questions on the basis of a newly developed, non-parametric method which uses response times and choice frequencies to distinguish genuine preferences from noise. We extend the method to allow for nontransitive choices, enabling us to identify the share of weak stochastic transitivity violations that is due to nontransitive preferences. By applying the method to two different datasets, we document that a sizeable proportion of transitivity violations reflect nontransitive preferences. Specifically, in the two datasets, 19% and 14% of all cycles of alternatives for which preferences are revealed involve genuinely nontransitive preferences. These violations cannot be accounted for by any noise or utility specification within the universe of random utility models.

We consider the aggregate and distributional impact of Universal Basic Income (UBI). We develop a model to study a wide range of UBI programs and financing schemes and to highlight the key mechanisms behind their impact. The most crucial channel is the rise in distortionary taxation (required to fund UBI) on labor force participation. Second in importance is the decline in self-insurance due to the insurance UBI provides, resulting in lower aggregate capital. Third, UBI creates a positive income effect lowering labor force participation. Alternative tax-transfer schemes mitigate the impact on labor force participation and the cost of UBI.

The share of low-income countries in global exports nearly tripled between 1990 and 2015, driven largely by the rapid emergence of China as an exporting powerhouse. While research in economics had long acknowledged that trade with lower-income countries could raise income inequality in Europe and the US, empirical estimates indicated only a modest contribution of trade to growing national skill premia. However, if workers are not highly mobile across firms, industries and locations, then the unequal impacts of trade can manifest along different margins. Recent evidence from countries across Europe and the US shows that growing import competition from China differentially reduced earnings and employment rates for workers in more trade-exposed industries, and for the residents of more trade-exposed geographic regions. These adverse impacts were often largest for lower-skilled individuals. We show that domestic manufacturing employment declined much more in countries that saw a large growth of net imports from China (such as the UK and the US), than in countries that maintained relatively balanced trade with China (such as Germany and Switzerland). Drawing on a new analysis for the UK, we further show that trade with China contributed to job loss in manufacturing, but also to substantial declines in consumer prices. However, while the adverse labour market impacts were concentrated on specific groups of workers and regions, the consumer benefits from trade were widely dispersed in the population, and appear similarly large for high-income and low-income households. Globalisation has thus created pockets of losers, and recent evidence indicates that in addition to financial losses, residents of regions with greater exposure to import competition also suffer from higher crime rates, a deterioration of health outcomes, and a dissolution of traditional family structures. We argue that new import tariffs such as those imposed by the US in 2018 and 2019 are unlikely to help the losers from globalisation. Instead, displaced workers may be better supported by a combination of transfers to avert financial hardship, skills training that facilitate reintegration into the labour market, and place-based policies that stimulate job creation in depressed locations

We extend the swaps index of rationality, introduced by Apesteguia and Ballester (2015) for a finite set of alternatives, to the standard consumer choice setting with infinite commodity spaces. Applications include consumer demand from competitive budget sets and the state-space approach to choice under uncertainty. We are interested in Apesteguia and Ballester’s result that the swaps index recovers the decision-maker’s true preference from choice data for a large class of boundedly rational behavioral models. We show that this result still holds in the consumer choice setting under a suitably defined monotonicity condition. This condition is satisfied for various models of interest but violated for others.

This paper studies random vectors X featuring symmetric distributions in that i) the order of the random variables in X does not affect its distribution, or ii) the distribution of X is symmetric at zero. We derive a number of characterization results for such random vectors, thereby connecting the distributional symmetry to various notions of how (Euclidean) functions have been regarded as symmetric. In addition, we present results about the marginals and conditionals of symmetrically distributed random vectors, and apply some of our results to various transformations of random vectors, e.g., to sums or products of random variables, or in context of a choice probability system known from economic models of discrete choice.

Automatic face recognition is a research area with high popularity. Many different face recognition algorithms have been proposed in the last thirty years of intensive research in the field. With the popularity of deep learning and its capability to solve a huge variety of different problems, face recognition researchers have concentrated effort on creating better models under this paradigm. From the year 2015, state-of-the-art face recognition has been rooted in deep learning models. Despite the availability of large-scale and diverse datasets for evaluating the performance of face recognition algorithms, many of the modern datasets just combine different factors that influence face recognition, such as face pose, occlusion, illumination, facial expression and image quality. When algorithms produce errors on these datasets, it is not clear which of the factors has caused this error and, hence, there is no guidance in which direction more research is required. This work is a followup from our previous works developed in 2014 and eventually published in 2016, showing the impact of various facial aspects on face recognition algorithms. By comparing the current state-of-the-art with the best systems from the past, we demonstrate that faces under strong occlusions, some types of illumination, and strong expressions are problems mastered by deep learning algorithms, whereas recognition with low-resolution images, extreme pose variations, and open-set recognition is still an open problem. To show this, we run a sequence of experiments using six different datasets and five different face recognition algorithms in an open-source and reproducible manner. We provide the source code to run all of our experiments, which is easily extensible so that utilizing your own deep network in our evaluation is just a few minutes away.

Since 1980, the world economy has experienced an increase of dominant firms. Dominant firms face limited competition in their market and exert monopoly power. Why has this happened, and why did it start in 1980? The rise of dominant firms has a direct impact on customers who pay higher prices, but it also has far-reaching implications for the macroeconomy. Widespread market power leads to wage stagnation and a decline in the labor share, it increases wage inequality, it slows down business dynamism, it reduces the number of startup firms and lowers innovation. In this public paper Eeckhout reviews the determinants of the rise of dominant firms, discusses the causes and consequences, and proposes directions for policy solutions.

In this paper I study whether citizens’ tax morale (and, more broadly, citizens’ attitudes towards the state) can be affected by past institutions, focusing on the role of historical fiscal capacity. Exploiting the features of the tax collection system of a pre-unification state in XIX Century Italy I identify differences in local historical fiscal capacity (as proxied by geographical proximity to a tax collector) and map them into contemporary tax morale, as measured by evasion of the TV Tax in 2014. Exploiting only variation in historical fiscal capacity that arises within matched pairs of neighbouring towns on the border of tax districts, I find imprecisely estimated and arguably small differences in tax morale today between towns where fiscal capacity was different. Investigating the mechanisms of transmission, I provide evidence that phenomena associated with structural transformation are likely to have halted the persistence of the historical fiscal capacity effect.

In a model inspired by neuroscience, we show that constrained optimal perception encodes lottery rewards using an S-shaped encoding function and over-samples low-probability events. The implications of this perception strategy for behavior depend on the decision-maker’s understanding of the risk. The strategy does not distort choice in the limit as perception frictions vanish when the decision-maker fully understands the decision problem. If, however, the decision-maker underrates the complexity of the decision problem, then risk attitudes reflect properties of the perception strategy even for vanishing perception frictions. The model explains adaptive risk attitudes and probability weighting as in prospect theory and, additionally, predicts that risk attitudes are strengthened by time pressure and attenuated by anticipation of large risks.

Return event studies generally involve several firms but there are also cases when only one firm is involved. This makes the relevant testing problems, abnormal return (AR) and cumulative abnormal return (CAR), more difficult since one cannot exploit the multitude of firms (by using a relevant central limit theorem, say) to design hypothesis tests. We propose a permutation test which is of nonparametric nature and more generally valid than the tests that have previously been proposed in the literature in this context. We address the question of the power of the test via a brief simulation study and also illustrate the method with two applications to real data.