The impact of tax increment financing on property value

• DOI: https://doi.org/10.1108/IJHMA-05-2019-0049
• Pages: 689-711
• Date: 19 September 2019
• Published date: 19 September 2019
• Author: Chaoran He,SeyedSoroosh Azizi
• Subject Matter: Property management & built environment

The impact of tax increment

f‌inancing on property value

Chaoran He

Department of Economics, Northern Illinois University, DeKalb, Illinois, USA, and

SeyedSoroosh Azizi

Purdue University Northwest, Hammond, Indiana, USA

Abstract

Purpose –Tax increment f‌inancing (TIF) has been adopted widely by municipalities to promote local

economicdevelopment. This study aims to examine the effect of TIF adoptionon property values at the parcel

level in Indianafrom 2009 to 2016.

Design/methodology/approach –Concerns of TIF adoptionendogeneity are addressed by a two-stage

estimationprocess using urban population ratio and unemploymentrates as instruments.

Findings –In addition to f‌inding inf‌luential socioeconomic and demographic factors, the results suggest

that parcels located within TIF districts were sold more than parcels outside of TIF districts by

approximately $5,000. Such premium is mainly picked up by the positive effect on commercial and

agriculturalparcels, which outweighs the negativeTIF impact on residential types.

Originality/value –Arm’s length transaction data on propertyvalue are used to eliminate the subjective

assessment bias, potentialcalculation errors during the evaluation process and econometricissues caused by

using the assessedvalue.

Keywords Housing, Housing policy, Housing markets, Property value, Housing taxation,

Tax increment f‌inancing

Paper type Research paper

1. Introduction

Allocating public resources to simultaneously meet the demand of economic development

and the vested interests of publicentities has been a central task for policymakers given that

misallocating such scarce fundsleads to high f‌iscal pressures on the local governments and

results in staggered local developments. Facing persistent f‌iscal constraints and decaying

infrastructure systems in recent years, local governments have long sought economic

development incentives (EDI) to combat the municipal deterioration and urban renewal

(Carroll, 2008). Tax increment f‌inancing (TIF), a type of EDI implemented by the local

jurisdictions, has grown into one of the most popular f‌inancing tools to cultivate local

economic development (Man and Rosentraub, 1998). Revenues generated within a TIF

district are set aside to defray the cost of thepublic infrastructure improvement within the

designated district (Weber and Goddeeris, 2007). With the success of TIF programs mainly

measured by property values,the intention of TIF is to increase future property tax revenue

of the local government by generatinga surplus from the initial value of the targeted land.

Although TIF aims to increase the property value, empirical evidence investigating the

TIF effect on property values is inconclusive. One possible cause is the estimations assume

that property values are equal to the assessment value (Anderson, 1990;Man and

Rosentraub, 1998;Dye and Merriman, 2000;Carroll and Eger, 2006;Byrne, 2006;Carroll,

2008; and Yadavalli and Landers, 2017);however, the assessed value can signif‌icantly differ

Tax increment

f‌inancing

689

Received14 May 2019

Revised21 June 2019

Accepted25 June 2019

InternationalJournal of Housing

Marketsand Analysis

Vol.13 No. 5, 2020

pp. 689-711

© Emerald Publishing Limited

1753-8270

DOI 10.1108/IJHMA-05-2019-0049

The current issue and full text archive of this journal is available on Emerald Insight at:

https://www.emerald.com/insight/1753-8270.htm

from arm’s length sales price in nature because of heterogeneity in attributes and

infrequency of sales (Rappaport, 2007). Housing is an extremely heterogeneous good that

varies greatly in physicalattributes among individual properties; and infrequent sales cause

prices of properties to be unobserved for unpredictable periods of time. The deviation in

value caused from using assessed value couldlead to a series of discrepancies in estimated

results when predictingTIF effect.

To eliminate the potential assessmentbias and related the econometric issues, this study

estimates the effect of TIFs by using arm’s length sales data on property valueat the parcel

level in Indiana from 2009 to 2016.Using urban population ratio and unemployment rates as

instruments, this study uses a two-stage probit maximum likelihood methodto address the

potential endogeneityof TIF implementation, and the results suggest that properties located

within the TIF district exhibit about $5,000 more in sales price, on average, than parcels

located outside of TIF districts. However, this overall positive effectmainly stems from the

commercial and agriculturalparcels whereas the TIF impact on residential propertiesturns

out to be negative.

The remaining structure of the paper is as follows. In Section 2, we summarizethe main

f‌indings of existing literature that examine the TIF effect on property values with special

focus on their econometric methods. Section 3 lays out the data and specif‌ications used in

the study, followed with a discussion on two sets of explanatory variables, including TIF-

adoption factors and property value determinants. Section 4 provides estimation results

from the two-stage probit maximum likelihood method, along with robustness checks with

alternative methods.Finally, conclusions are discussed in Section 5.

2. Literature review

The primary question that researchers have attempted to answer in the related literature is

whether property values inside of TIF districts are higher than those outside.

Notwithstanding that TIF is explicitly designed to promote local developments, the

empirical f‌indings on the general effect of TIF on the property value nationwide remain

mixed, with the discrepancy possibly stemmingfrom the unobserved factors impacting the

decision to designatea TIF district or different data available for variousanalyses.

Anderson (1990) is the f‌irst, to the best of the authors’knowledge,to raise the concern of

the potential reverse causality between TIF adoption and property value growth because

municipalities could deliberatelyadopt TIF districts in areas with high property value for a

higher diverted revenue. By modeling growth rates of aggregated property values of 255

Michigan cities, he addresses the endogeneitywith a two-stage probit maximum likelihood.

He f‌inds that TIF adopting cities experiencedgreater property value increases than non-TIF

adopting cities in Michigan from1985 to 1986. Note that the property value is aggregated at

the city level and measuredby 1986 State Equalized Valuation.

Man and Rosentraub (1998) examine the effect of TIF on the property value growth of

151 cities in Indiana from 1970 to 1992, with a similar two-stage model to address the

causality. The f‌irst stage estimates a probit maximum likelihood regression to predict the

probability of an area being designated as a TIF district, with the per capita federal aid,

property tax revenues and tax rates as instruments; and the second stage uses a f‌irst-

difference regression between the post- and pre-TIF property values. By adding the

computed selectivity variable from the f‌irst stage as a regressor in the second stage, the

f‌irst-difference regression results suggest that the property value, measured by median

owner-occupied housing value (taken from a database complied by Indiana Board of Tax

Commissioners), in TIF-adopting cities was 11 per cent greater than in the cities without

TIF.

IJHMA

13,5

690