HomeMy WebLinkAboutTax Increment Financing Memo
~
~HILLS
MEMORANDUM
DATE: November 26, 2008
TO: Honorable Mayor and City Council
FROM: Ron Moorse, City Administrator
SUBJECT: "Tax Increment Financing: It's Effect on Local Government Finances"
Attached is a copy of the University of Minnesota, Center for Urban and Regional Affairs CURA
REPORTER, Volume 33 featuring "Tax Increment Financing: Its Effect on Local Government Finances".
This publication was provided by Council Member Dave McClung.
UNIVERSITY OF MINNESOTA
Center for Urban and
Regional Affairs
PORTER
V OLU!\1E 33
NU:tvfBER 2
SUIvTh-fER 2003
Tax Increment Financing:
Its Effect on local Government Finances
by Kenneth A4. Kriz
Tax increment financing (TIF) is a
widely used economic develop-
ment and urban redevelopment
tool. Local governlnents use TIF to
finance im.provement projects by
floating a general bond that is paid off
using the additional tax revenue gener-
ated by the improvement project.
Although TIF allows municipalities to
creatively finance development projects
and can be an effective means of
promoting development in blighted
areas, public perception of TIF has
always been somewhat tenuous. Son1.e
have suggested that TIF diverts resources
that could be better used elsewhere.
Others argue that developers use TIF
and other economic development
incentives to "playoff" cities against
one another and make additional
profits, demanding incentives even
though they would have moved their
operations to an area anyway. Finally,
some question the effect of TIF on local
finances, suggesting that this redevelop-
ment approach may promote increased
demand for public services and cause
local governments to lose out on the
growth in tax base that would have
occurred without TIE
Many practitioners and researchers
have struggled with the question of how
TIF affects local government finances.
The prevailing wisdoln usually takes one
of two forms: the "pure capture" theory
and the "pure attribution" theory. The
pure capture theory views TIF as simply
a nleans to redirect public resources to
subsidize developn1.ent. Proponents of
this view argue that all development
would likely occur without the use of
\:J
::J
o
S
0-
'<
~
ro
<
ro
V')
t"'l
::J
:J
ro
0:
~
In This Issue:
. Tax Increment Financing: Its Effect on Local Government Finances .........1
[J Project Update: Statewide Digital Parcel Map Inventory ....................8
. Perceptions of the Environmental Review Process in Minnesota . . . . . . . . . . . .9
o 19th Annual Conference on Policy Analysis .........................., 3
[J 2000 Population Change Maps ...................................1 3
. The Community Assistantship Program: Connecting the University with
Greater Minnesota ...................................................'6
[J Project Awards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .22
o Project Update: Bringing INFO-U to New Audiences . . . . . . . . . . . . . . . . . . . . . .23
public subsidy. Therefore, any resources
expended on TIF are a pure subsidy to
developers. The pure attribution theory
sees TIF as a necessary incentive to lure
businesses to areas that would likely
never be developed, at least for a period
of several years. Any resources going to
TIF are merely investments that will
likely payoff handsomely in the future.
My research takes a different
approach to this topic. Using a proba-
bility-based analysis technique called
Monte Carlo simulation, I investigated
the probabilities of a positive financial
outcome fronl TIE By sim.ulating nlY
model under various scenarios, I was
able to answer the following important
questions concerning TIF:
1. What is the likely effect of TIF proj-
ects on local government financial
conditions?
2. Are TIF projects good investments
for local governm.ents (Le., are they
likely to produce rates of return that
justify their use)?
3. U.nder what conditions are the best
financial results from TIF obtained?
This study was jointly funded by the
Minneapolis Planning Department and
by a grant froIn CURA's l~aculty Interac-
tive Research Program. With this
support and with the assistance of a
graduate research assistant provided
through CURA, I .was able to compile an
annotated bibliography of TIF resource
materials for use by the Minneapolis
Planning Department, collect and
analyze property tax records from
Hennepin County and the Minneapolis
Community Development Agency, and
develop a deterministic simulation
model that the city of Minneapolis can
use to assess the financial effects of
proposed TIF projects.
This article will provide a brief
explanation of how TIF works in theory
and practice, describe the nlethodology
for the simulation Inodel used to assess
the financial effects of TIF projects,
discuss the results of this siInulation,
and offer conclusions and policy impli-
cations stemming fronl this research.
What Is Tax Increment Financing?
How TIP Works. 'I'he nature of tax
increment financing varies from state to
state. In Minnesota, a local government
Cover photo: Calhoun Square, located
in the Uptown neighborhood of
Minneapolis, was constructed in the
1980s using tax increment financing.
2 CURA REPORTER
or its designated redevelopment
authority determines that an area of (or a
specific property in) the community is
econonlically challenged due to some
factor or set of factors-although often
this determination emerges out of a
request by a developer for a subsidy. 'The
local government or other redevelop-
ment authority then goes through a
process to designate a tax increment
district (TID) that includes the area
targeted for redevelopnlent (in
Minnesota there are seven types of
districts that can use TIF including rede-
velopment districts and mined under-
ground space districts). The assessed
value of the TID is then "frozen." This
means that all future revenues from
properties in the TID are split between
two different destinations. Future tax
revenues derived from the value of the
frozen assessed value continue to go to
the various local governments as they
did before the I'll) was designated.
However, taxes derived from increases in
the assessed value above that frozen
value go to the authority in charge of the
TID to pay for development costs (the
additional revenue is the tax increment).
These development costs can include
physical development of propelties as
well as disposition of the properties at
below market prices, or write-downs. 1
Table 1 and the sidebar on page 4
briefly document the history of the use
of TIF in Minnesota.
The Economic Impact of TIF. Tax
increment financing affects the pattern
of both revenue and expenditures for
local governments. On the revenue side,
Tn~ mostly inlpacts the growth rates of
assessed market value for parcels of
property in and out of the 'TII), as well
as altering the number of properties
froln which the local government
derives revenue for basic city services.
On the expenditure side, the increase in
economic wealth in the jurisdiction
(reflected in the increase in assessed
market value) is likely to cause increases
in demand for public goods, thereby
raising expenditure demands.
In the theoretical model that was
developed for this research project, the
variables with the most effect on the
finances of local government were the
following:
.... the pre- and postproject growth rates
of properties located inside or
outside the tax increment district
1 In SOIne if not most states, TID authorities can
issue bonds, the principal and interest on which
are repaid from the tax increment. .
.... the probability that the properties in
a TIF district would have been devel-
oped without the use of TIF
.... the elasticity of demand for expendi-
tures on local government services
with respect to changes in assessed
valuation
.... the direct operating subsidy provided
by the local government to a TIP
district
Study Methodology
The uncertainty inherent in the deci-
sion to use TIF mitigates against the use
of simple models to analyze financial
effects. It is easy to see that if a property
is likely to be developed without the use
of TIl~ the revenue effect for the city is
negative. The relevant policy question
is, What is the amount of the revenue
effect, on average? If it is likely that
revenue will be greatly reduced through
using TIF, cities may want to consider
other means of financing projects, or
perhaps not offer the projects full local
government support. However, if the
implementation of TIF will likely cause
only a slnallloss of revenue or even
revenue gains (through increases in the
market value of other properties in the
jurisdiction), local governments should
seek all possible opportunities to use
this redevelopment tool.
Three factors make it difficult to
predict the fiscal inlpacts of TIE First,
little research had been done on the
effects of TIP on economic development.
'There have been only a handful of
attempts to address the question of rela-
tive econolnic growth rate and the
results of these studies are split, with
some finding a positive effect on
economic growth and others finding no
effect or even a negative effect. Second, it
is not possible to objectively determine
the probability of future development
without TIF before the decision is made to
use TIE Finally, it is possible only to
estirnate future growth rates of properties
and the future elasticity of demand for
public services. Therefore, these are
sources of uncertainty for potential
policy prescriptions regarding TIE
In order to address these issues, I
used a Monte Carlo simulation. Monte
Carlo simulations were first developed
in the 1940s during the Manhattan
Project and are currently used in fields
ranging from quantum physics to
finance. The simulations allow an
analyst to predict potential future values
for a theoretical model by entering a
large number of random inputs and
then recording outputs generated from
Table 1. Tax Increment Financing District Data for Taxes Payable 1974-2002
Source: Minnesota Revenue Department, Property Tax Unit, 2003
a Number of different cities and towns with tax increment districts.
b Number of different counties that contain tax increment financing districts.
C Total value of the cities and towns with tax increment districts.
d Value used. in determining increment district taxes. It is after the deduction of any captured value shared with the local taxing districts and after the deduction of any
captured value contributed to the fiscal disparity pool in the seven-county metropolitan area.
e Ratio of retained captured value to total value.
f Tax generated by multiplying the local tax rate by the retained captured value. Beginning in 1989, gross tax increment includes local tax rate excess tax increment,
which is distributed entirely to the local taxing districts.
9 Assessed values for taxes payable 1974-1988. Tax capacity values for taxes payable 1989 and thereafter.
SUMMER 2003 3
the model. The outputs-which are
usually in the forIn of probability distri-
butions-show.the range of potential
risks and rewards of decisions based on
the theoreticalluodel under study.
In conducting a Monte Carlo simula-
tion for the fiscal impacts of TIF, I created
a deterministic model of the 'TIF'decision-
making process. I then replaced uncertain
variables in the model-the growth rate
of non-TIF properties, the effect of TIF on
non-TIF properties, the post-TIP growth
rate of TIF properties, and the elasticity of
expenditures with respect to changes in
4 CURA REPORTER
market value-with random variables
drawn from a specified distribution. The
uncertain variables and the sources from
which the random variables were derived
are listed in Table 2.
Along with the random variables,
certain parameters were established
wi thin the luodel. Parameters are vari-
abies whose values are more certain or
more controllable than the uncertain
variables. For example, if one were
simulating traffic flow on a freeway, one
would set as a parameter the number of
lanes on the freeway. One can vary
parameters, but usually this is done only
to analyze policy changes. Initially, the
following parameters were set: initial
investment into the TIP project, effec-
tive tax rate, operating subsidy, pre- TIF
growth rate of TIF properties, and
discount rate. The initial investment
parameter was set at 500/0 of the TIF
market value. In other words, we
assume that a $30 million TIF project
will generate a $15 million initial
market value investment. The city effec-
tive tax rate was set at 0.919'6, calculated
from City of Minneapolis financial
Table 2. Data Sources for Random Variables
* City of Minneapolis, Comprehensive Annual Financial Report.
t Joyce Man and Mark Rosentraub, "Tax Increment Financing: Municipal Adoption and Effects on Property
Value Growth," Public Finance Review 26 (1998): 523-547.
* Richard Dye and David Merriman, "The Effects of Tax Increment Financing on Economic Development,"
journal of Urban Economics 47 (2000): 306-328.
~ Minneapolis Community Development Agency, Comprehensive Annual Financial Report.
records.2 Finally, the operating subsidy
was set at zero. Operating subsidies vary
widely by project and usually have the
characteristics of a loan, repaid out of
future tax increlnent. It should be
noted, however, that this assumption
likely biases the result of this analysis
somewhat in the direction of showing a
n10re positive financial effect from TIE
The pre-TIF growth rate of the prop-
erties that will eventually be part of the
TID can be treated either as a paralneter
or as a policy variable (that is, a variable
that can be changed easily by a govern-
111ental entity). I treated the growth rate
as a paralneter and assumed a zero
growth rate. (For states such as
Minnesota that allow the base to
increase by an amount equal to the pre-
TIF growth rate, this factor should drop
out.) Another variable that can be
treated as either a policy variable or a
2 City of Minneapolis, C01Jzprehensive Annual Finan-
cial Report. The tax rate is an effective one, found
by dividing property tax revenue by total market
value for the most recent available year.
Table 3. Monte Carlo Simulation Base
Results (10,000 iterations)
parameter is the discount rate. Again,
without any prior knowledge of when
proj ects will be developed, I treated this
as a parameter and assumed a 70/0
discount rate.
The final variable in the model is the
probability that a parcel will develop
without the use of TIE I chose to treat
this variable as a policy variable. Given
that TIF allows local governments to
tenlporarily take properties off local tax
rolls, an important question is, which
properties should be incorporated into
l'II)s? For Iny initial estimates, I
assumed a 100/0 probability of develop-
ment. This translates into one TIF prop-
erty being developed without 'rII~~ 011
average, within the next 10 years.
Results
The model outlined in the previous
section was simulated through 10,000
iterations using the @Risk add-in
package for Microsoft Excel. Table 3 and
Figure 1 show the results. Table 3 shows
descriptive statistics and the distribution
of the results. The first column of the
table shows the net present value for
the project as a whole evaluated at the
discount rate of 7<Yo specified above.
This column shows that the most likely
result (point estimate) of the effect of
this $15 million project (with $30
million captured market value) is
slightly more than a $5.5 million net
financial loss to the local government.
(Figure 1 shows that the distribution of
results is skewed, therefore it is appro-
priate to cite the median value as the
point estimate.) The mean is statistically
significantly greater than its standard
error (p < .001), meaning that we can
say with a high level of confidence that
there is a negative financial effect for
the local government. Table 3 and
Figure 1 also present an estimate of the
Figure 1. Probability Distribution of Net Present Value, Base Estimates
(10,000 iterations)
450/0
400;u
35%)
30<*,
>, 250/0
u
~
OJ
;::s
0-
OJ
;.... 20eyo
~
15%)
10eyo
5 <Yo
00/0
-25 -22.5 -20 -17.5 -15 -12.5 -10 -7.5 -5 -2.5 0 2.5
7.5 10 12.5 15 17.5 20 22.5 25
Net Present Value (in millions of dollars)
SUMMER 2003 5
Table 4. Simulation Results Using Various Probabilities of Development without the Use of TIF
distribution of the potential risks and
rewards of using TIP. For exalnple, Table
3 shows that there is a 50/0 chance of a
loss greater than $9.4 million and a gain
greater than $3.3 million. There is
slightly more than a 900/0 chance of a
negative net present value. Figure 1
presents a histogranl of the results for a
project's net present value. The distribu-
tion has a high peak and is positively
skewed, with the greatest probability
falling around the mean and the vast
majority of results falling within a range
fron1 -$10 million to +$2.5 Inillion.
One of the questions that motivated
this study was, under which conditions
would the local governlnent stand the
greatest chance of reaping a positive
financial impact from TIF? My theoret-
ical model indicates that one of the
most inlportant variables in deter-
mining the financial effects of TIF is the
probability that the parcel will be devel-
oped without the use of TIP. For the
second siInulation, I varied the proba-
bility of developlnent. It appears that
overoptimisnl about the financial effects
of 'I'IF should be a larger concern among
policy makers. Table 4 shows the shnu-
lation results assuming various probabil-
ities of development. There are at least
two striking results from these simula-
tions. First, even if there is a zero proba-
bility of development without the use of
TIF, the nledian net present value is
negative (approximately -$2.7 million).
Even when there is no chance of future
development, there is only a 30(Yo
chance of a positive financial outcome
for the local governnlent. Increases in
the probability of development up to
10% produce additional expected finan-
ciallosses. However, another interesting
result is that the point estinlate of the
loss converges toward a stable amount
with probabilities of development
6 CURA REPORTER
greater than 109'6. This means that after
the likelihood of development exceeds a
certain amount, the negative financial
effect becomes somewhat constant.
Therefore, the most likely range of
losses on a $15 million TIF project is
somewhere between $2.5 million and
$6.5 million.3
Another area of potential policy
importance is the length of time that
the TIF district is allowed to capture
incremental revenue from the properties
in the TID. As stated earlier, redevelop-
ment districts in Minnesota that use TIF
are allowed to capture increment for a
period up to 25 years. However, in other
states (and for other types of districts in
Minnesota), the time limits differ. The
impact of this policy decision on local
government financial condition is
shown in Table 5. Table 5 presents the
results for the base case of a 25-year TIF
capture and an alternative 10-year TIP
capture. The mean and median results
are significantly greater for the 10-year
capture limit. This suggests that time
limits might be an effective tool for
easing the financial burden of TIF
inlposed on local governlnen ts.
Conclusions
This study illuminates at least four
important points about the effect of TIF
on local governnlent finances. Pirst,
given a realistic set of policy assump-
tions, TIP most likely produces a net
financial loss to a local governn1.ent.
Only when the probability of develop-
ment without the use of TIF is low and
the positive effects of TIP on property
values both inside and outside of the
'TIP district become extremely high can
3 The point net present value estimate at a 500/0
probability of development is -$5.4 rnillion, which
is not statistically significant at accepted levels
(p::: .06).
a TIP project be justified in financial
terms.
Second, although the first conclu-
sion holds, there is nonetheless a
nonzero probability of a positive finan-
cial impact. A single point estimate
simply does not capture enough infor-
mation for local decision makers to
form an opinion about the financial
effects of TIP. If events work in its favor,
the local government can expect a small
financial gain from TIE In short, TIP
may be best thought of as a large finan-
cial investment by local governments
1:7
::r
o
8"
~
~
([)
<
([)
V)
("\
::r
:J
([)
a.:
~
Burnsville~ Heart of the City project, a
mixed-use commercial and residential
development on Nicollet Avenue, is
being constructed through the use of
TIF funds.
Table 5. Simulation Results for Different Tax Increment District (TID) Certification limits
that lnay have a sm.all chance of finan-
cial reward.
Third, even if a TIF project produces
net financial gains in the long run, in
the near term TIF will likely cause nega-
tive financial effects. Although projects
with positive expected net present
values should be implenlented, a policy
of widespread use of TIF may be unsus-
tainable from a financial standpoint.
The ultiInate question for elected local
officials is whether 'TIF proj ects expected
to have long-term positive financial
effects can be impleInented in the face
of other short-term budget dem.ands.
Those seeking to justify the use of TIF
on the basis of nonfinancial rewards
should be prepared to address how these
rewards might offset the financial losses
froIn the use of the particular incentive.
Finally, we can now say a bit more
about those circumstances that will lead
to a profitable project from the local
government's perspective. These circum-
stances include low probability of devel-
oplnent without TIF and a short
certification period.
As this study demonstrates, local
governments face significant financial
risks when they inlplement TIF projects,
but that they can potentially reap finan-
cial rewards. Currently TIF is an open
policy option for conlmunity and
economic developnlent in most states,
and it will likely remain so for many
years to come. However, much further
research is necessary to provide context
on the decision to use TIF.
Epilogue: Is There a Future for TIF in
Minnesota?
During the 2001 legislative session (after
the bulk of the research for this article
was complete), the Minnesota State
Legislature proposed and the governor
signed a bill changing the way that
primary education is funded in the state.
The principal effect of the bill was to
"remove" the basic general education
levy from the local government property
tax. An unintended consequence of this
change was to relllove one significant
financial participant from tax increment
distlicts (before the changes, school
district levies constituted an average 340/0
of the local property tax levy).
Many observers have speculated that
this change has struck the death knell
for TIF in Minnesota. Previously, cities
and counties wanting to use TIF shared
the burden of financing TIF with school
districts, each providing about one-third
of the financial value. Now with the
majority of school district finances
conling from sources other than the
local property tax, cities and counties
will have to participate with nearly all
of the necessary revenues for the
project. Other economic development
tools-such as the property tax
abatement incentive authorized by the
Minnesota legislature in 1997
(Minnesota Statutes 2001, ~469.1813)
-may be more attractive for local
governments.
Education finance reform may well
have struck a blow to the use of TIF in
Minnesota, but there are at least a
couple of ways that local governments
can still employ TIE First, a local
government could simply enlarge the
district that uses the financing tool. A
district with a larger geographic area
would allow a jurisdiction to capture
more resources to use for the project.
Second, although property taxes are
currently the only authorized source of
tax increment funding in Minnesota, it
is neither inconceivable nor unprece-
dented for local governments to use
other tax increments to finance
econom.ic redevelopment.4 For the
moment, TIF in Minnesota lllay not be
dead, but only in hiatus.
Kenneth A. Kriz is assistant professor of
public finance at the School of Public
Administration, University of Nebraska at
Omaha. He was assistant professor of
public and nonprofit management at the
Hubert H. Humphrey Institute of Public
Affairs at the time the research for this
article was undertaken. His current
research focuses on municipal debt
management, economic development
policy, and transportation finance, along
with the use of alternative estimation
techniques in public finance. He teaches
courses in public sector economics, public
finance, and statistical analysis.
The research upon which this article is
based was supported in part through a
New Initiative grant from CURA. These
grants support projects that are initiated
by faculty, community organizations,
government agencies, or students and
that are not appropriate for consideration
under another CURA program. For full
results of this research project, see
Kenneth A. Kriz, "The Effect of Tax Incre-
ment Finance on local Government Finan-
cial Condition." Municipal Finance Journal
22 (2001): 41-64.
4 For a review of other tax increment prograrns, see
John L.Mikesell, uNonproperty Tax Increlnent
Progranls for Econonlic Developrnent: A Review of
the Alternative Programs." In Tax Increment
Financing and Economic Development: Uses, Stnlctures,
and l1npact. Craig L. Johnson and Joyce Y. Man, eds.
Albany, NY: SUNY Press, 2001.
SUMMER 2003 7