Josef Montag

Criminals and the Price System: Evidence from Czech Metal Thieves

People steal copper and other nonferrous metals to sell them to scrap yard. Simultaneously, prices at scrap yards are set by the world market. We argue that shocks in metal prices represent a quasi-experimental variation in gains from crime. This allows us to estimate the behavioral parameters of supply of offenses and test the economic theory of criminal behavior. Our estimates suggest that the long-term elasticity of supply of metal thefts with respect to the re-sale value of stolen metal is between unity and 1.5. Moreover, the system tends to equilibriate quickly—between 30 and 60 percent of a disequilibrium is corrected the following month and the monthly price elasticity estimates are around unity.

Homeownership, Mobility, and Unemployment: Evidence from Housing Privatization

Homeownership is believed to cause higher unemployment. This is because homeowners face higher mobility costs that limit their job search to local labor markets. Empirical tests of this prediction have yielded mixed results so far, possibly due to the endogeneity of homeownership. This paper documents that the privatization of public housing in Central and Eastern Europe after the fall of the Iron Curtain resulted in a quasi-experimental assignment of homeownership to individual households. This facilitates a new test of the effects of homeownership on mobility and unemployment. We find only weak evidence that homeowners are less willing to move and no evidence of higher unemployment risks relative to renters.

Is Pro-Labor Law Pro-Women? Evidence from India

I study the effects of state-level differences in labor regulation on labor market outcomes of women in India. Using a representative sample of urban households from 2005, I find that labor regulation has a large negative effect on women’s economic activity, mainly employment. My estimates suggest that a one standard deviation increase in the labor regulation measure decreases the probability of a woman being economically active by 3% to 4%—the implied decrease in female labor force is between 15% and 18%. The effects on men’s participation are around zero. I do not find labor regulation to have a significant effect on male wages or on the gender wage gap. Finally, labor regulation is associated with women having less say at home and a lower sex ratio.

Legal Origins and Labor Market Outcomes of Men and Women

The paper exploits variation in institutional environment and regulation of labor across legal origins to explain international differences in gender-gender income ratio, income inequality, labor force participation, unemployment, and selfemployment. Relative to common law countries, women have higher relative incomes in French (16%), Scandinavian (14%), and Post-socialist (15%) countries, but not so i German law countries. In Scandinavian and Post-socialist countries, female labor-force participation is about 15% higher relative to common law countries, whereas in French law countries, women exhibit lower participation and tend to be more often unemployed (4%) or self-employed (6%). Women are also more often self-employed in German law countries. A plausible interpretation is that in strong (past) active labor market policies may be driving results in Scandinavian and Post-socialist countries, whereas labor regulation in French law countries burdens marginal workers while their observed relative wages are pushed up.

Let the Punishment Fit the Criminal: An Experimental Study

We use a laboratory experiment to study the extent to which people tailor levels of punishment to the subjective experience of the person to receive that punishment, for both monetary and non-monetary sanctions. We find that subjects tend to apply higher fines to wealthier individuals. Additionally, subjects assign more repetitions of a tedious task to those with a lower willingness-to-pay to avoid it. We find no evidence that the distributions of monetary and non-monetary punishments are different when considered as proportions of the maximum possible punishment, but that this does not hold when non-monetary punishments are converted into monetary equivalents. This suggests that subjects do not have in mind a particular level of disutility from the punishment, but rather are guided by the sentencing possibilities.

The Effects of a Simplified Criminal Procedure: Evidence from One Million Czech Cases

We estimate the effects of a simplified criminal procedure applicable to minor crimes on case durations and probabilities of charges and conviction. The identification strategy exploits a quasi-natural experiment in the implementation of the simplified procedure across districts. The procedure reduces the duration of the pre-trial phase and increases the probability that the prosecutor will charge the defendant in court. The effects on the duration of the court phase and the probability of conviction at trial are less significant. The resources released by the use of the simplified procedure could be allocated to serious cases. However, we do not find evidence of such beneficial spillovers.

The Effects of a Simpler Criminal Procedure: Evidence from One Million Czech Cases

The paper estimates the effects of a simpler criminal procedure on case durations and the probabilities that the defendant is charged and convicted. The identification strategy exploits a quasi-natural experiment in district-level implementation of criminal procedure reform in the Czech Republic. The reform allowed petty offenses to be prosecuted via a simplified (fast-track) procedure. We find very strong evidence that prosecuting a case via the fast-track procedure reduces the duration of the police/prosecutor phase of the procedure and increases the probability that the prosecutor charges the suspect at court. To a lesser extent, it also reduces the duration of the court phase of the procedure and increases the probability of conviction at trial. The fast-track procedure released resources that could potentially be spent on prosecuting serious cases. We investigate for possible spillover effects but find no evidence of their presence. Our estimates suggest that specialization of prosecutors and court senates decreased after the reform, possibly mitigating indirect efficiency gains from the reform.

Identifying Odometer Fraud: Evidence from the Used Car Market in the Czech Republic

This paper investigates the presence of odometer fraud in the used-car market in the Czech Republic using a unique dataset of 250,000 car-sale ads. Alternative identification techniques are also discussed. However, selection into the market as well as the practice of rounding odometer readings---possibly strategic yet innocent---render the standard statistical tests unusable. A modification of the last-digit test, which was previously used to detect fraud in election and accounting data, is therefore developed and employed. The results suggest that suspicious patterns are more prevalent in the segment of cars imported from abroad. I also show that this methodology can be used at the firm-level, which may be of interest to authorities and market participants.