Do State Corporate Tax Incentives Create Jobs? Quasi-experimental Evidence from the Entertainment Industry

• Published date: 01 June 2019
• DOI: 10.1177/0160323X19877232
• Date: 01 June 2019
• Author: Michael Thom
• Subject Matter: General Interests

General Interest
State and Local Government Review
2019, Vol. 51(2) 92-103
Do State Corporate Tax
ª The Author(s) 2019
Article reuse guidelines:
Incentives Create Jobs?
sagepub.com/journals-permissions
DOI: 10.1177/0160323X19877232
Quasi-experimental Evidence
journals.sagepub.com/home/slg
from the Entertainment
Industry
Michael Thom1
Abstract
Policy makers allocate billions of dollars each year to tax incentives that increasingly favor creative
industries. This study scrutinizes that approach by examining motion picture incentive programs
used in over thirty states to encourage film and television production. It uses a quasi-experimental
strategy to determine whether those programs have contributed to employment growth. Results
mostly show no statistically significant effects. Results also indicate that domestic employment is
unaffected by competing incentives offered outside the United States. These findings are robust to
several alternative models and should lead policy makers to question the wisdom of targeted
incentives conferred on creative industries.
Keywords
economic development, tax, tax incentive
For over a century, state and local policy mak-
of targeted incentives shows no sign of abating.
ers have sought to encourage economic devel-
Sixty-eight percent of state and local govern-
opment by offering incentives that target
ments offered them in 1999; by 2009, it was
specific firms and industries. But targeted
95 percent (Florida 2018). The roots of that
incentives have only recently drawn consider-
growth lay in the political environment. Policy
able scrutiny, thanks in part to their escalating
makers use incentives to signal proactiveness
cost. For example, Tesla agreed in 2014 to build
on the economy, and targeting a specific firm
a factory in Nevada after officials there offered
or industry brings greater visibility to their
tax and other incentives valued at US$1.3
billion. Foxconn decided in 2017 to locate new
facilities in Wisconsin in response to incentives
1 Price School of Public Policy, University of Southern
valued at between US$3 billion and US$4.5
California, Los Angeles, CA, USA
billion. Several governments later competed
over Amazon’s HQ2 project with incentive
Corresponding Author:
Michael Thom, Price School of Public Policy, University
packages worth as much as US$8.5 billion.
of Southern California, 650 Childs Way, MC 0626,
Whether through so-called megadeals or
Los Angeles, CA 90089, USA.
other programs that attract less notice, the use
Email: mdthom@usc.edu

---

Thom
93
efforts than offering nonparticularized incen-
Targeted Economic Development
tives. If the target ultimately locates in an
Incentives in Context
area that proposed incentives, policy makers
benefit by taking credit. If it settles else-
The inclination toward targeted economic
development approaches in the United States
where, policy makers may still benefit by
has roots in the Great Depression. During that
taking credit for making an attempt to create
period, policy makers in southern states
jobs, deflecting any blame to problems
enacted bond programs that subsidized facto-
beyond their control.
ries and other facilities, thereby lowering firms’
Of course, policy makers do not act in a
effective capital costs. That bond-supported
vacuum. Offered a choice between a policy
infrastructure was publicly owned and exempt
maker who offers incentives and one who
from property taxation yielded further cost
does not, voters prefer the former even if
advantages (LeRoy 2005). Many observers
both attract the same development (Jensen
believed this tactic successfully enticed labor
and Malesky 2018). For their part, busi-
and capital from northern states, where policy
nesses develop rent-seeking relationships
makers responded with retaliatory incentives.
with policy makers to protect incentives
Competition accelerated through the 1970s and
against electoral turnover—a quid pro quo
1980s and became more global in scope (Jenn
that metastasizes to other rent-seeking
and Nourzad 1996). Incentives evolved toward
arrangements (Coyne, Sobel, and Dove
further particularization as policy makers—
2010; McChesney 1997).
seeking ever-narrowing competitive advan-
If there’s a voice of caution in the milieu, it
tages—began to target specific firms and indus-
comes from those who assess targeted incen-
tries as well as locations (e.g., downtown cores,
tives. Indeed, studies commonly find that they
enterprise zones, and brownfield sites) and
do not yield promised benefits (e.g., Peters and
events (e.g., the Summer Olympics).
Fisher 2004). But evaluative research has failed
The accumulated findings of an extensive
to keep pace with targeted incentives’ prolifera-
literature on targeted incentives converge
tion, making it difficult for policy makers to
toward a single conclusion about their efficacy.
judge whether or not they’re a prudent use of
In short, studies suggest policy makers should
resources.
avoid the practice altogether, especially with
This study investigates the employment
incentives that carry tax expenditures, because
impact of motion picture incentive (MPI)
they fail to stimulate commensurate economic
programs, a combination of corporate tax
gains (e.g., Fox and Murray 2004; Hicks and
incentives and other services made available
LaFaive 2011; Kolko and Neumark 2010; Neu-
by over thirty state governments to encour-
mark and Kolko 2010; Patrick 2014; Reese
age film and television production. MPI pro-
2014). In instances where gains materialize,
grams are one component of a broader
they may be short term (e.g., O’Keefe 2004;
strategy across those governments to diver-
Wassmer 1994; see also Hamersma 2008).
sify economies by incenting a creative indus-
Lackluster outcomes have many causes.
try believed to yield stable, high-wage jobs.
Most state and local business tax frameworks
To that end, policy makers in some states
are not appreciably different, rendering any sin-
have approved higher tax expenditures for
gle incentive unable to rouse substantial firm
MPI programs than many prominent mega-
relocation or expansion (Wasylenko 1997).
deals. As such, they are a relevant case from
Furthermore, taxes for many industries are not
which to draw implications about the effi-
a primary operating cost. Among manufactur-
cacy of targeting an industry with exclusive
ers, for example, taxes compose around 1 per-
incentives—in this case, a creative industry
cent of input costs compared to over 21
with a high degree of mobility and, in theory,
percent for labor (Keynon, Langley, and Paquin
high sensitivity to those incentives.
2012). Marginal tax reductions offered in one

---

94
State and Local Government Review 51(2)
area may thus fail to compensate for other costs
expanded those services and added corporate
that may be higher in the same area. Competi-
tax incentives that were not available to other
tive targeting can also create a zero-sum game
sectors (Christopherson and Rightor 2010).
in which one area’s “win” comes at the expense
These MPI programs eventually spread to
of another’s loss (Chirinko and Wilson 2008;
forty-four states, carried by rising unemploy-
Wilson 2009).
ment and domestic competition (Leiser 2017;
Targeting nevertheless endures, and the tar-
Thom and An 2017).
gets have evolved. Policy makers in many state
Although the number of MPI programs has
and local governments have oriented their eco-
declined, investment has not. In 2017, accord-
nomic development strategies toward growing
ing to state government reports, over thirty
the “creative class” and building “creative
states granted the industry a combined
cities” into a “creative economy” (Florida
US$1.7 billion in corporate income tax expen-
2002; Howkins 2001; Scott 2000). The strategy
ditures, not including the value of other pro-
is vested in a belief that creative, knowledge-
gram services. About 77 percent was
intensive industries—that is, sectors in which
concentrated in five high-expenditure states
intellectual property is the output—produce
(New York, Louisiana, Georgia, Connecticut,
stable, high-wage jobs that serve as a growth
and Massachusetts) that represented only 58
driver and a buffer against economic shocks.
percent of expenditures five years earlier.
The argument is especially attractive in areas
Cumulative spending in these states rival those
where policy makers have struggled to revive
for prominent economic development mega-
economies decimated by losses in manufactur-
deals (see Table 1).
ing and other heavy industries.
Each high-expenditure state’s MPI program
But the approach engenders an intractable
has common features, including location assis-
Catch-22. For all the benefits of a creative
tance, advertising, and preferential regulatory
economy, there are drawbacks, among them
treatment. Some include sales and transient
gentrification, rising housing costs, and higher
occupancy tax waivers and incentives for
inequality. Compared to traditional industry
building production-related infrastructure.
clusters, the market for creative labor and cap-
States differentiate themselves with corporate
ital is more global and competitive (Florida
income tax credits that vary from 10 to 40 per-
2005). And relative to other industries (e.g.,
cent of production...