Drawing the Line: Felony Theft Thresholds and the Strategic Valuation of Larceny

Abstract

This thesis examines whether felony larceny thresholds induce strategic bunching in reported theft values. Felony thresholds are monetary cutoffs which distinguish a misdemeanor charge from a felony charge based on the value of property stolen. These thresholds create a discrete cutoff in which a marginal increase in the value of stolen property can lead to a substantial increase in the penalty imposed. Using incident-level data from the National Incident-Based Reporting System (NIBRS), the analysis first tests whether theft values cluster around felony thresholds. Then, the paper investigates whether this clustering shifts to the new threshold when there is a change. In particular, Virginia will be utilized as a case study having raised its felony threshold from $200 to $500 in 2018, and from $500 to $1,000 in 2020. To do so, this thesis combines a bunching framework with difference-in-differences and triple-difference estimations. The analysis finds that the distribution of theft values is discontinuous at the felony threshold, though part of this pattern is driven by heaping in the data. The regression results reveal that when Virginia increased its threshold, theft became more concentrated around the new cutoff, with stronger effects above the threshold rather than below it. This finding suggests that felony thresholds function as salient price signals, and that the distribution of theft values is also influenced by administrative valuation practices rather than by offender behavior alone. As an extension, the paper examines Fairfax County’s 2020 prosecutorial policy in which thefts under $1,500, rather than the statutory $1,000 threshold, would not be charged as felonies. The results indicate that this “prosecutorial threshold” also functions as a salient signal and induces bunching around the cutoff.