Does Tax Increment Financing Pass the “But-for” Test in Missouri?

Abstract

The use of tax increment financing (TIF) remains a popular, yet highly controversial, tool among policy makers in their efforts to promote economic development. This study conducts a comprehensive assessment of the effectiveness of Missouri’s TIF program, specifically in Kansas City and St. Louis, in creating economic opportunities. We build a time-series data set starting 1990 through 2012 of detailed employment levels, establishment counts, and sales at the census block-group level to run a set of difference-in-differences with matching estimates for the impact of TIF at the local level. Although we analyze the impact of TIF on a wide set of indicators and across various industry sectors, we find no conclusive evidence that the TIF program in either city has a causal impact on key economic development indicators.

Did incentives help municipalities recover from the Great Recession? Evidence from Midwestern cities

“…Furthermore, we cannot replicate the approaches of some researchers (e.g., El-Khattabi & Lester, 2017;Lester, 2014) who track the implementation of incentives and economic outcomes over time in order to support causal inferences and bolster attribution claims. First, in compiling our incentives dataset, we were unable to obtain data specifying the instigation and trajectory of the incentives and so cannot discern the history of each instance of incentive use.…”

“…One approach to ruling out systematic differences that might influence outcomes is to identify a control group of firms that did not receive incentives but is similar in other ways to firms that received them. El-Khattabi and Lester (2017) discovered that Census block groups in TIF districts in Kansas City, Missouri fared no better over time in terms of growth in employment, sales, or the number of establishments than matched locations lacking the TIF incentive. 2 Instead, evaluations of the outcomes of economic development programs tend to default to the places where subsidized businesses operate.…”

“…For example, Gabe and Kraybill (2002) find that Ohio businesses receiving state economic development incentives experienced slower growth than non-subsidized businesses. El-Khattabi and Lester (2017) attribute the inferior economic performance of St. Louis locations within TIF districts to development trajectories that differ from non-incentivized areas of the city and possible pretreatment bias in the designation of TIF locations. In both of these studies, the authors control for potential selection bias in TIF designation and firm selection (i.e., that slower growing areas are more likely to adopt TIF in the first place).…”

“…(In their regression models, however, only wages were significant.) El-Khattabi and Lester (2017) discovered that Census block groups in TIF districts in Kansas City, Missouri fared no better over time in terms of growth in employment, sales, or the number of establishments than matched locations lacking the TIF incentive. In St. Louis, locations within TIF districts actually fared worse than their non-TIF counterparts; the authors attribute the counterintuitive finding to selection bias, since TIF designations may be targeted intentionally toward locations with the poorest economic prospects.…”

“…In both of these studies, the authors control for potential selection bias in TIF designation and firm selection (i.e., that slower growing areas are more likely to adopt TIF in the first place). El-Khattabi and Lester (2017) attribute the inferior economic performance of St. Louis locations within TIF districts to development trajectories that differ from non-incentivized areas of the city and possible pretreatment bias in the designation of TIF locations. One exception is Man (1999), who estimates a cross-sectional, time series regression for Indiana municipalities, finding that municipal adoption of TIF exerted a significant positive effect on local employment.…”