Andrew Jordan

I am an Assistant Professor of Economics at Washington University in St Louis.  I apply the theoretical and empirical tools of economics to the criminal justice system.

Email: awjordan@wustl.edu

Link to CV

Active Research:

This paper develops a model of plea bargaining with risk averse defendants. Defendants vary in risk aversion, which is private information. Prosecutors make plea offers that trade sentence length off against the risk of a costly trial. I structurally estimate the model using data on serious felony cases from Chicago. This estimation shows that defendants are exceptionally risk averse, plea sentences significantly exceed expected trial sentences, and average defendant surplus from plea bargaining is positive but small. I also examine differences by race, finding that black defendants are more tolerant of risk. (Formerly What Can Plea Bargaining Teach Us About Racial Bias in Criminal Justice?)


Prosecutors have unchecked power to drop charges, which may generate racial disparities. Race-specific unconditional conviction rates, prerequisite to analysis of these disparities, are estimated with a novel bounding argument relying on random assignment of cases. Data from the Felony Review Unit in Cook County show a modest tendency to approve charges against non-Black suspects. Identified sets for unconditional conviction rates overlap, but cases against Black suspects are typically less likely to end in conviction. Cook County does not currently do felony review in narcotics cases, but including them can improve both equity and efficiency.


We study how civilian complaint investigators affect officer behavior in Chicago. We exploit quasi-random assignment of complaints to supervising investigators and use variation in whether supervisors tend to acquire sworn affidavits that substantiate the complaints. When the assigned investigator opens more investigations through obtaining affidavits, accused officers accumulate fewer complaints in the first three months of the investigation. We find that, prior to a scandal, assignment to high-investigation supervisors causes officers to make more arrests. However, this reverses after the scandal. Our findings suggest that police watchdogs can improve officer behavior in ordinary oversight environments but may backfire in heightened oversight environments. [This paper previously circulated as "The Effects of Investigating Police Misconduct"]


This paper studies the effects of prosecutorial charging decisions in Cook County. Felony narcotics cases cannot ordinarily be reduced to misdemeanors but can when combined with other charges. In such cases, misdemeanors rise by 5.3 to 7.1 percentage points. Defendants shift over both the no charges-misdemeanor margin and the misdemeanor-felony margin. The combined effect of these shifts is a modest fall in recidivsm. Relative to others, Black defendants are more likely both to be on the felony-misdemeanor margin and to reduce recidivism. These results approximate a policy that permits misdemeanor narcotics charges without disallowing felony charges


One oft-suggested policy to improve equity in policing involves increasing racial diversity among police leadership, but there is little evidence on how this would affect outcomes.  We show, in the context of the Chicago Police Department (CPD), that police teams managed by Black and Hispanic supervisors make fewer low-level arrests.  To overcome potentially endogenous absences, we exploit the pre-determined, rotational nature of the CPD operations calendar and compare within unit-watch outcomes on days when Black and Hispanic police lieutenants (LTs) are expected on duty to (otherwise similar) days when they have a day off and another LT is in charge.


We examine ten cohorts of eighth grade students in public schools in Chicago, IL: 1995-2004. We focus on male students and find that composite measures of math achievement, reading achievement, and neigh-borhood SES during elementary school are strong predictors of future felony arraignment and incarceration rates, even among students of the same race who attend the same school. Nonetheless, our measures of elementary achievement and early SES account for less than half of Black versus non-Black disparities in arraignment and incarceration rates. Results derived using value-added measures of eighth grade quality suggest that schools may reduce criminal justice involvement, particularly among Black males, by better preparing students for the non-cognitive demands of high school.

Published and Forthcoming:

We examine 70,581 felony court cases filed in Chicago, IL during the period 1990-2007. We exploit case randomization to assess the impact of judge assignment and sentencing decisions on the arrival rates of new charges. Our estimates of the impact of incarceration on recidivism show that, in marginal cases, incarceration creates large and lasting reductions in recidivism among first offenders. yet, among repeat offenders, incarceration sentences for marginal offenders create only modest short-run incapacitation effects and no lasting reductions in the incidence of new felony charges. Our results raise concerns about sentencing reforms, enacted in most states over recent decades, that encourage of mandate incarceration sentences for many offenders with prior criminal records. 

Using a compiled dataset of 441 censuses and surveys between 1787 and 2015, representing 103 countries and 48.4 million mothers, we find that: (1) the effect of fertility on labor supply is typically indistinguishable from zero at low levels of development and large and negative at higher levels of development; (2) the negative gradient is stable across historical and contemporary data; and (3) the results are robust to identification strategies, model specification, and data construction and scaling. Our results are consistent with changes in the sectoral and occupational structure of female jobs and a standard labor-leisure model.