Article Review

N a tio n a l Tax J o u rn a l, S e p te m b e r2 0 1 4 , 6 7 (3), 6 7 5 - 6 9 6

TA X IN C R E M E N T D E B T F IN A N C E A N D T H E G R E A T R E C E S S IO N

Martin J. Luby and Tima Moldogaziev

This paper explores U.S. local government debt finance activities related to Tax Increment Financing (T1F) between 2000 and 2013. We gather comprehensive data about debt that is serviced through TIF, document changes in several variables related to the amount, use, and structural features o f such debt, and evaluate the impact o f the Great Recession on these variables. Our results indicate that the Great Recession limited how local governments could sell and structure TIF debt. We suggest that these limitations were the result o f the limited capital available during and immediately after thefinancial crisis, structural changes in the financial industry caused by the financial crisis, and increased risk aversion by investors.

Keywords: Tax Increment Financing (TIF), property tax, economic development, Great Recession

JEL Codes: H71

I. INTRODUCTION

By design, tax increment financing (TIF) involves a delay between the redevelopment costs paid and the project benefits received. Because o f the mismatch between the time o f incurring development costs and the receipt o f benefits o f greater property or sales tax revenues associated with economic development, TIF naturally lends itself to the use o f debt finance. Over the last several decades, many local governments have used debt instruments to securitize future tax revenues in order to pay for current costs related to the economic development o f the TIF district (Johnson, 1999). Such securitization often involves the selling o f securities in the U.S. municipal bond market. However, the financial crisis o f 2007-2008 greatly affected the use and structure o f the municipal securities market (Johnson, Luby, and Moldogaziev, 2014). In addition, the financial crisis deflated the value o f one o f the primary repayment pledges o f these TIF securities

M a rtin J. Luby: School o f P ublic Service, DePaul U niversity, and In s titu te fo r G o v e rn m e n t and P ublic Affairs, U niversity o f Illinois, C hicago, IL, USA (m lu b y l @ depau!.edu)

T im a M o ld o g a z ie v : S chool o f P u b lic & In te rn a tio n a l Affairs, U n iv e rs ity o f G eorgia, A th e n s, GA, USA (tim atm @ u g a .ed u )

676 National Tax Journal

in most areas o f the United States, namely the local property tax base. For example, the city o f Louisville, Kentucky recently nearly defaulted on its 2006 TIF bonds sold to finance the KFC Yum! Center. TIF property tax revenues in 2012 amounted to $3.5 million with the expectation just a few years earlier that the TIF district would generate $8.2 million to pay the principal and interest payments on the bonds in 2012 (Boyd, 2013). It is likely that the Great Recession’s impact on property values significantly contributed to the shortfall between expected and realized TIF revenues in Louisville.

Given the extensive use o f municipal securities to finance TIF capital projects over the last several decades and the changes in the municipal securities market and local property tax base as a result o f the recent financial crisis, this paper explores the use of TIF debt finance before and after the Great Recession (i.e., during the period 2000-2013). We gather data about debt that is serviced through TIF and document changes in sev­ eral variables related to the amount, use and structural features of such debt, focusing our analysis on the impact the Great Recession had on these variables. We also offer a brief discussion on the possible future o f TIF debt finance in the post-Great Recession world.

II. TIF DEBT FINANCE MECHANICS

Before presenting data on the entire TIF industry from 2000-2013, this section o f the paper briefly details the purpose and mechanics o f TIF debt finance. Local governments generally use TIF debt finance, which often entails the sale of municipal securities, for three reasons (Johnson, 1999). First, as mentioned above, the sale o f municipal securi­ ties allows the local government to quickly raise a large amount o f financial resources for TIF redevelopment projects. In the absence o f municipal bonds, the local govern­ ment would have to rely on loans from banks and/or developers or to significantly decelerate the pace o f the redevelopment projects to be in line with annual increases in the tax base. In addition, municipal securities are generally tax-exempt so the local government receives the benefit o f lower cost financing with municipal securities vis- a-vis bank or developer loans. Second, TIF debt allows municipalities to circumvent constitutional or statutory debt restrictions, since TIF bonds are generally not subject to general obligation bond debt limitations or public referendum requirements (Johnson, 1999; Briffault, 2010). Thus, local governments can access redevelopment resources without seeking legislative and/or public approval. Finally, TIF provides local govern­ ments an opportunity to raise off-balance sheet capital financing since the issuer o f TIF bonds, generally a redevelopment agency, is usually not considered part of the general government. This serves to preserve the local government’s borrowing capacity for future capital projects. In essence, the basic pros and cons o f pay-as-you-go financing versus pay-as-you-use financing are present in the decision to use TIF debt finance.

Local governments sell TIF bonds in the national municipal bond market. The prevail­ ing feature o f this market is tax exemption. That is, the interest on municipal securities is generally exempt from federal income taxation. However, federal tax regulations specify that in order to qualify for such tax exemption, at least 95 percent o f the TIF bond proceeds must be used for redevelopment purposes in a “blighted” area, the issuer

T a x I n c r e m e n t D e b t F in a n c e a n d t h e G r e a t R e c e s s io n 677

must have a redevelopment plan, and the pledge o f repayment must be from general taxes o f the government or from incremental taxes associated with the project (Johnson, 1999). If these conditions are not met, the local government will incur higher interest costs as it will have to sell the debt on a taxable rather than tax-exempt basis. As it relates to amortization structure, local governments sell the debt with a bond structure that approximates the expected size o f the future incremental tax revenues associated with the redevelopment. That is, the local government amortizes the TIF bonds over a time period during which incremental revenues are expected to be sufficient to pay the principal and interest payments on the debt.

In the parlance o f the municipal bond market, TIF bonds essentially represent a hybrid general obligation/revenue bond credit structure (Geheb, 2009). TIF debt appears to be general obligation in nature in that ad-valorem property taxes often secure repayment on the bonds. However, TIF debt also carries a revenue bond feature in that there is a specific, identifiable revenue repayment source (i.e., the incremental tax revenues) that would not exist in the absence o f the redevelopment project. Moreover, redevelopment agencies, rather than the general government, are often the issuer o f the TIF debt just as other types o f government authorities are often issuers of revenue bonds. This hybrid structure has raised some constitutional issues related to the level o f responsibility that the general government has in “making whole” TIF bond investors in the event o f a default (Geheb, 2009). In the past, local governments have sold much (although not all) o f their TIF debt using a revenue bond structure (Johnson, 1999).

III. T H E T IF M U N IC IPA L SECURITIES MARKET, 2 0 0 0 -2 0 1 3

In this section o f the paper we discuss the data — the trends, use, and structure of TIF debt securities issued between January 1,2000 and December 3 1 ,2 0 1 3 — with an emphasis on the impact o f the Great Recession. This covers several years before and after the financial crisis of 2007-2008, including the years before and after the Great Recession. The database includes a list o f all TIF debt securities sold during this time period ( N= 2,478) as well as a spectrum o f features o f these bond transactions. The data for this descriptive analysis come from the Securities Data Corporation (SDC).' This is a fee-for-service database company that collects information on all municipal securities sales. Our review considers TIF market trends, the issuers and uses o f TIF securities, and the structure o f TIF securities. As such, the analysis provides a complete picture of the TIF municipal securities market over the last 14 years for the period January 2000 through December 2013. We assess the distribution o f TIF securities issuance data for goodness-of-fit using Pearson’s and log-likelihood Chi-squared tests o f independence, which test the likelihood o f whether the observed annual distributions for variables of interest in this study are due to chance or whether there are significant shifts in their annual distributions during the period under examination. The tests compare the observed distributions in the data to the expected distributions based on the assumption that the variables are independent. While these tests do not provide evidence o f the direction o f

These data are available at http://thomsonreuters.com/sdc-platinum/.

678 N ational Tax Journal

association between the variables o f interest, in our case, they provide evidence on the probability o f independence between each o f our variables o f interest and annual TIF bond issuance activity in 39 U.S. states and the District o f Columbia.

A. TIF M arket Trends

In the period, almost $37.6 billion in 2,478 separate TIF issues were sold in the municipal bond market. As seen Figure 1, the least TIF issuance, both in terms o f monetary volume and the number o f individual issues, was in calendar year 2012. Only 65 TIF issues for a combined volume o f $449 million were recorded in the municipal market during that year. The largest volume o f TIF securities is observed in 2006 with almost $5 billion sold in 238 individual issues. Overall trends show that TIF issuance

Figure 1

TIF Issuance Volume (SMillion) and Frequency in 39 U.S. States and the District o f Columbia, 2000-2013

6,000 300

5,000 250

= 4,000 200

3,000 150

~ 2,000 100

1,000 50

Year

SSSS! Frequency of TIF Issues ■ Grand Total

Source: Securities D ata C orpo ra tio n

T a x I n c r e m e n t D e b t F in a n c e a n d t h e G r e a t R e c e s s io n 679

volume grew from under $2 billion in 2000 and 2001, to about $2.7 billion in 2002, to more than $4.8 billion in 2003, before declining to about $3.4 billion in 2004. In the three years immediately before the Great Recession, years 2005, 2006, and 2007, TIF issuance remained significantly above $4 billion. We find that about 75 percent of all TIF-related securities, more than $28 billion, were issued before the Great Recession.

TIF volume, however, dropped dramatically during and after the Great Recession. Thus, in 2008 TIF issuance volume fell almost by half to about $2.1 billion from the pre-recession year’s level o f $4.2 billion. TIF volumes further fell to about $1.5 billion in 2009, climbed back to $ 1.8 and $2 billion in 2010 and 2011, respectively, but shrunk to less than half a billion in 2012. The TIF volume in 2013 appears to be about $1.5 billion. Only $9.5 billion was issued in TIF securities since the Great Recession from 2008-2013. Consequently, it appears that there was a significant decrease in TIF debt activity in the municipal securities market since the Great Recession. The volume o f TIF bond sales decreased by 49 percent between 2007 and 2008 with every subsequent year lower than the 2008 level, which suggests the detrimental impact the financial crisis may have had on the size o f the TIF bond market. Indeed, the measures o f association between the number o f TIF bond issues and the year o f issuance provide evidence of statistically significant changes in issuance trends in our sample period. For the period o f January 1, 2000 to December 31, 2013, Pearson’s Chi-squared and log-likelihood Chi-squared statistics are 818.1 and 730.1, respectively.

The State o f California has been the “market leader” in TIF securities, except in 2012. The dramatic decline in 2012 was a result of California’s dissolution of its redevelopment authorities in late 2011, which prohibited California local governments from issuing any new TIF bonds (Lefcoe and Swenson, 2014). California was a pioneer in the use o f TIF debt finance and was perennially the largest seller o f TIF bonds prior to the elimination in its redevelopment agencies. The steep drop between 2011 and 2012 in the volume and number of TIF transactions reflects this statutory change. In 2013 California returned as the largest issuer o f TIF securities with 21 separate TIF issues for a combined value of $618 million. However, these municipal securities represented refinancing o f existing TIF securities, which were allowed under Assembly Bill 1484. Refinancing of outstand­ ing TIF securities was allowable as long as the amount o f the refinancing bond issue was not greater than the amount o f refinanced bonds, and the refinancing bond interest costs were less than refinanced bond interest costs.2 In the period under study, California municipalities are directly responsible for over $25 billion in TIF securities, which is roughly about two-thirds o f the entire TIF activity in the municipal securities market. A total o f 59 municipal issuers from California appear to have delivered at least $ 100 mil­ lion o f TIF securities each to the market, with the San Jose City Redevelopment Agency leading the way with almost $2.5 billion in 20 separate TIF issues.

The data indicate a similar impact o f the Great Recession on overall TIF debt finance activity even if California is removed from the database. As shown in Figure 2, the largest

2 County of Los Angeles Redevelopment Refunding Authority, Tax Allocation Revenue Refunding Bonds, Series 2013DEF, Official Statement, available at http://emma.msrb.org/ER733654-ER569382-ER970689.pdf.

680 N ational Tax Journal

F ig u re 2

TIF Issuance V olum e ($M illion) and Frequency 38 U.S. States (Excluding California) and th e D istrict o f C olum bia, 2000-2013

1,800

1,600

~ 1,400 C0 1 1,200

CD § 1,000

8 800 c CO

I 600 LL H 400

200

0

Year

Frequency of TIF Issues ■Grand Total

180

160

140

120 | =3 c r <D

100 £ 0)o

80 | Uiu>

60 tfc I-

40

20

0

S o u rce : S e c u ritie s D a ta C o r p o r a tio n

and smallest volumes o f non-Califomia TIF issuance were in 2005 and 2012, respectively, with a sharp decline in TIF transactions and volume between 2007 and 2008 and a steady decline thereafter with an uptick in transaction frequency in 2013. Even after omitting California issuers, the measures o f association between the number o f TIF bond issues and the year o f issuance provide evidence o f statistically significant changes in issuance trends in our sample period. For the period from 2000-2013, Pearson’s Chi-squared and log-likelihood Chi-squared statistics are 658.0 and 600.8, respectively. Aside from California, there are five states that have issued more than $ 1 billion each in TIF securi­ ties during the period under review. These states are Colorado, Missouri, Minnesota, Illinois, and Texas. In Figure 3, we depict aggregate annual volumes in these top-five TIF issuing states for 2000-2013. Though overall Colorado leads the top-five list with about $1.6 billion, its municipalities are the least frequent market participants with only 41 separate TIF issues. Its biggest aggregate TIF issues o f $296 million and $266 million

Tax Increm ent Debt Finance and the Great Recession 681

682 N a tio n a l T a x J o u r n a l

are in 2004 and 2008, respectively. On the other hand, Minnesota municipalities are the most frequent TIF issuers with 410 separate TIF securities for a combined volume of $1.4 billion. Texas, Illinois, and Missouri appear to fall in between Colorado and Min­ nesota in terms o f their transaction frequencies in the TIF market.

B. Is s u e rs a n d U se s o f T IF S e c u r itie s

Not coincidentally, issuers from California dominate the TIF market in the period. O f the 10 largest issuers o f TIF securities (a combined volume o f over $8 billion or over 20 percent o f the entire industry in the period), eight are in California. By far the largest issuer o f TIFs in the US was the San Jose Redevelopment Agency with almost $2.5 billion in TIF securities (Table 1). The San Francisco City & County Redevelop­ ment Agency issued over $1.1 billion during the same period. Three remaining major issuers in California were the Oakland Redevelopment Agency, the San Diego Rede­ velopment Agency, and the Riverside County Public Finance Authority — all with TIF volumes exceeding $600 million. These top-5 issuers in California accounted for about 8 percent o f all TIFs in the state in 2001 and about 40 percent in 2009, with their shares in other years falling somewhere between these extremes. Nevertheless, despite the continued activity in the TIF market by San Francisco City & County-, Oakland-, and San Diego Redevelopment Agencies since the Great Recession, it is evident that most o f the TIF activity for these top-5 California issuers occurred before the Great Recession.

As shown in Table 2, apart from California jurisdictions, the Denver Urban Renewal Authority is the largest issuer o f TIFs by volume in our sample. It issued more than $800 million in TIFs (or more than half o f all TIF securities issued by Colorado municipalities), with $171 million sold as recently as 2013. Atlanta and Chicago are next with $657 million (more than 96 percent o f TIF securities in Georgia) and $476 million (over 37 percent o f the entire TIF issues in Illinois), respectively. Minneapolis and the Unified Government o f Wyandotte County & Kansas City have sold $344 mil­ lion and $270 million in TIF securities each. These latter two issuers accounted for a significant fraction o f TIF securities from their states; about 26 percent for Minneapolis and over 76 percent for Wyandotte County/Kansas City. The Unified Government of Wyandotte County & Kansas City has not returned to the TIF market since 2005, how­ ever. When combined, these top-5 non-California issuers o f TIF securities accounted for about 2 percent o f all non-California issues in 2012, but about 40 percent in 2000.

We classify the over thirty categories o f TIF debt uses identified in the data into seven broad categories. As Table 3 shows, general purpose/public improvements are by far the biggest share o f the TIF market in our sample with about $20 billion in proceeds (or about 53 percent o f all TIF volumes). Almost 80 percent o f general purpose/public improvements TIF securities were issued before the Great Recession, however. Eco­ nomic development and industrial development TIFs are next in volume. Over $13.5 billion were issued for these uses as TIF obligations. TIF proceeds for these three major

T a x I n c r e m e n t D e b t F in a n c e a n d t h e G r e a t R e c e s s io n 683

684 National Tax Journal

Tax Increm ent D ebt Finance and the Great Recession 685

686 N a tio n a l Tax J o u rn a l

uses accounted on average for 89 percent o f all TIF securities in the m unicipal market. The TIF m arket has thus been predom inantly a m arket for general purpose/public im provem ent and econom ic and industrial developm ent projects.

The remaining annual volumes (on average roughly 11 percent o f all TIF debt proceeds) are used for transportation, housing, education and health, and utilities, w ith housing, both single and multifamily structures, adding up to more than $2 billion. Transportation uses, such as overland infrastructure, airports, and seaports, accounted for more than $1.6 billion o f TIF proceeds. Education TIFs, and to a lesser extent healthcare related uses, absorbed about $330 m illion o f debt proceeds. A very small fraction o f the TIF m arket ($ 184 m illion in our sample) is related to water, sewer, and gas purposes, electric and public pow er purposes, solid waste and recycling uses, as well as hybrids o f these.

The economic development category generally shows a steady decline in issuance volume for the period from 2000 to the Great Recession period. Increases in the general purpose/public improvement category initially offset this decline, but eventually decline between 2007 and 2008. After 2008, the economic development category increases while the general purpose/public improvement category continues to fall. W hile these category names are somewhat generic (especially “general purpose/public improvement”) and one must use caution when drawing conclusions, the decline in economic development TIF bonds and the increase in general purpose/public improvement TIF bonds suggests that local governments no longer use TIF only for economic development but also to finance general governmental purposes. The uptick in economic development bonds after 2009 and the decline in general purpose bonds may be the result o f demands by municipal bond investors for more specific details on the development projects being financed rather than loaning money for “general purpose projects” that are financed by incremental tax revenues. However, this conclusion is speculative and warrants further systematic analysis.

A nother w ay to explore the use o f TIF proceeds is to look at trends in the tax status o f TIF securities. As stipulated in federal tax statute, typically at least 95 percent o f the use o f proceeds m ust be pledged for redevelopm ent purposes in a “blighted” area to attain tax exemption. Thus, analyzing the com position o f bonds sold tax-exem pt versus taxable m ay provide m ore insight into the purposes that local governm ents have used TIF debt, that is, w hether local governments have expanded their use o f TIF debt finance tools to fund redevelopm ent in non-blighted areas or not. Figure 4 illustrates the volume and percentage o f TIF bonds sold on a taxable basis. For m ost years betw een 2000 and 2014, local governm ents consistently sold their TIF debt on a tax-exem pt basis about 80 percent o f the tim e w ith the heaviest use o f taxable TIFs in 2010, 2011, and 2013. However, the overall volum es o f taxable bonds dropped dram atically since the Great Recession (i.e., taxable bond volum es in each year post-recession w ere less than h a lf o f the immediately pre-recession year levels). Private activity TIFs generally are sponsored by corporate sponsors and are alm ost always taxable securities. However, during the tim e period analyzed, only $347 m illion o f TIF issues (less than 1 percent o f the entire TIF market) were directly linked to corporate supporters as seen in Figure 5. M ost o f these were tied to pre-G reat R ecession years when private funds w ere still relatively abundant. Our data show that the frequencies o f issues for taxable and corporate-backed TIFs have decreased significantly, w ith P earson’s Chi-squared and log-likelihood Chi- squared statistics at 57.0 and 56.6 com pared to 17.8 and 17.6 respectively.

T ax I n c r e m e n t D e b t F in a n c e a n d t h e G r e a t R e c e s s io n 687

C. Structure of TIF Securities

The overall volume o f TIFs has decreased since the Great Recession and so did average bond issue sizes. Mean bond sizes peaked in 2005-2007 and decreased significantly by 2012 as reported in Figure 6. However, Figure 6 shows a larger average bond size in 2013. Another typical measure o f bond structure, length o f maturity, is generally positively related to risk, with investors demanding a higher risk premium for longer term investments or generally deciding to avoid longer-term investments altogether in extreme circumstances. In this context, one might expect to see the average final maturity decline in the years after the Great Recession as TIF bonds became riskier in light of the financial crisis and its effect on property values. The results in Figure 7 show a decline in 2009, an increase back to pre-recession levels in 2010 and 2011, and then a significant decline in 2012 and 2013. However, one needs to be careful in drawing any definitive conclusions from final maturity data without looking at times-to-call measures. As the measure o f average years to call in

Figures 4 -7 Percentage o f Total TIF Issuance Volume and Select TIF Features,

39 US States and the District o f Columbia, 2000-2013

30

25

P 20 CD o 15

10o<D CD 8 5

001

Figure 4 Taxable TIFs

i 1 :

ilI I O CM ^ CD CO OO O O O O t- O O O O O O CM CM CM CM CM CM

Figure 5 Corporate Backed TIFs 120

C O 100

80 <D N

CO 60 0 =3 CD (f)

40

LL H

20

0 _ 1 - hi__

Figure 6 Average TIF Bond Issue Size

o CM CD 0 0 O CMo O O O Oo O O O O O O CM CM CM CM CM CM CM

Figure 7 Average TIF Bond Final Maturity

O CM M - CD CO O CM O O O O O O O O O O O O CM CM CM CM CM CM CM

688 N a t io n a l T a x J o u r n a l

Figure 8 suggests, there was indeed a shortening o f time-to-call that would fit such ex- ante evaluations o f the market risk by the issuers o f TIF securities.

The coupon interest rate sets the periodic interest payments on local governments’ TIF bonds. In higher interest rate environments, coupon interest rates are generally higher and in lower interest rate environments, they are generally lower. However, the coupon interest rate is not the same as the yield that investors will receive (issuers will pay) because it does not take into account call provisions or the price of the securities. In Figure 9 the average highest coupon increases after 2007, and moves downward in 2012. This is especially interesting because interest rates generally declined after 2007 as a result o f the financial crisis, as the Federal Reserve Bank cut its benchmark interest rate several times and continued to maintain low interest rates well into 2014. If there was not an interest rate risk premium that investors were building into TIF bond inter­ est rates, we would expect to see the average highest coupon rate decline after 2007 in line with the general decline in interest rates. In fact, the average highest coupon rate is materially larger in the four years after the financial crisis (2008-2011) than the three years previous to the financial crisis when interest rates were generally higher. This lends some support to the notion that TIF bond investors may have viewed such securities as relatively risky and thus demanded higher interest rates after the Great Recession.

Many local governments sell their TIF debt as revenue bonds with repayment solely payable from the incremental tax revenues generated in the TIF district (Johnson, 1999). However, some local governments sell TIF debt using a general obligation bond structure whereby repayment is ultimately backed by the full faith and credit of the general taxing body. Due to this more robust repayment pledge, investors generally view general obligation bonds as less risky than revenue bonds. One might expect that local governments would increase the sale o f TIF debt using a general obligation bond structure rather than a revenue bond structure in the years during and after the Great Recession. In Figure 10, we observe that immediately before the Great Recession the share o f general obligation TIF securities was between 2 and 5 percent, and the by early 2000s it was between 7 and 11 percent. Since the financial crisis, however, the share of general obligation TIFs increased to as high as 27 percent. It appears that TIF issuers are relying on their full faith and credit in more debt issues after the recession than on issues before the recession. We see statistically significant changes in the number o f general obligations TIFs in the market, with corresponding Pearson’s Chi-squared and log-likelihood Chi-squared statistics at 53.8 and 53.4. O f course by pledging their general obligation credit for TIFs, municipal issuers are reducing their debt issuing capacity for non-TIF issues.

The method by which local governments sell their bonds — negotiated or competi­ tive — is an indicator o f whether local governments are using financial intermediaries to resolve information asymmetries. In a negotiated bond sale, a local government works directly with a pre-selected underwriter to pre-market and market their bonds tailoring the bond structure to market conditions and investor demands. In a competi­ tive sale, local governments sell their bonds to an underwriter or group o f underwriters via an open auction with the lowest bidder receiving the bonds. Among other things,

T a x I n c r e m e n t D e b t F in a n c e a n d t h e G r e a t R e c e s s io n 689

previous research has shown that the additional certification provided by underwriters in a negotiated bond sale can be advantageous (i.e., it can lead to lower interest costs compared to competitive sales) when there is significant volatility in the bond markets (Leonard, 1994; Kriz, 2003; Peng and Brucato, 2003). The financial crisis o f 2007-2008 created more volatility in the capital markets than in any other period in decades. With such volatility, all else equal, we might expect an upward spike in negotiated sales of TIF bonds. The TIF market was dominated by negotiated sales even before the crisis, but significantly more deals relied on negotiated sales since then. As shown in Figure 11, since 2009 at least 87 percent o f TIF securities have been sold in negotiated deals, except for the anomalous year 2012 when California withdrew from the TIF market. This shift is statistically significant as the Pearson’s Chi-squared and log-likelihood Chi-squared statistics are equal to 53.9 and 52.9.

Figures 8 -1 1

Percentage of Total TIF Issuance Volume and Select TIF Features, 39 US States and the District o f Columbia, 2000-2013 (c o n tin u e d )

F ig u r e 8 A v e ra g e Y e a rs to C a ll

O CM ^ CD CO O O O O O O O O O O O O CM CM CM CM CM CM CM

F ig u r e 9 A v e ra g e H ig h e s t C o u p o n

Y e

a rs

t o

C a

ll

O

C D

C

O

O

N >

^ 8 . 0 0 c 0

® 6 . 0 0

I 001f co § 2 . 0 0

0 . 0 0 O CM CD CO OO O O O O t- o o o o o o CM CM CM CM CM CM

0 F ig u r e 10 G e n e r a l O b lig a t io n T I F s o | 30 (/) 2 2 5 LL

H 20 CD o 15

o 10 CD O) d CD O

O CM NT CD CO O CM O O O O O O O O O O O O CM CM CM CM CM CM CM

I n i i . i m r

CD=3(/)0)

0 OH

M— o 0 O) CD

0 Q_

F ig u r e 11 N e g o t ia t e d T IF s

O CM CD 00 O CM O O O O O O O O O O O O CM CM CM CM CM CM CM

690 N ational Tax Journal

Prior research has also shown that the use o f financial advisors on a municipal security transaction can help deal with information asymmetries. These financial intermediaries help certify the true value o f the municipal security and thus reduce the borrowing costs of issuers (Johnson, 1994; Vijayakumar and Daniels, 2006). Based on this certification theory, we would expect the use o f financial advisors on municipal TIF securities to increase after the 2007-2008 financial crisis in order to alleviate the information asymmetries that increased at this time. Figure 12 shows the percent of total TIF bonds that were sold without the use o f a financial advisor. The share o f TIF bonds issued without a financial advisor declined significantly between 2007 and 2009 and stayed low throughout the entire post Great Recession period, as compared to the much higher shares observed in the pre-recession period. It appears the local govern­ ments sought out this additional market certification more frequently after 2007-2008 in light of the uncertainty created by the financial crisis. This conclusion is supported by highly significant Pearson’s Chi-squared and log-likelihood Chi-squared statistics of 78.6 and 84.2.

With more volatility in the capital markets and the collapse o f credit enhancement tools for TIF finance, better credit quality became important in the post-recession years. Figure 13 tracks the percentage o f TIF bond issues rated by at least one o f the three major credit rating agencies (Moody’s, Standard and Poor’s, or Fitch) from 2000-2012. Through their assignment o f credit ratings, these agencies are financial intermediaries that attempt to relieve the information asymmetries between the issuer o f bonds and investors especially with respect to the likelihood o f default. Local governments seek credit ratings to relieve these informational asymmetries as a means o f reducing their borrowing costs. Our results show that there was a general upward trend in the extent

Figures 1 2 -1 3

Percentage o f Total TIF Issuance V olum e and Select TIF Features, 39 US States and th e D istrict o f C olum bia, 2000-20 13 (continued)

00 § 45 1 40 LJL 35 H 30 3 25 t 20 ° 15 2 10 m 5

Figure 12 TIFs without FA

0 Q_ O CM ^ CO CO O CMO O O O O t- t-

o o o o o o o CM CM CM CM CM CM CM

Figure 13 TIFs without Rating

o CM CD 00 O CM o O O O O o O O O O O O CM CM CM CM CM CM CM

Tax In c re m e n t D e b t F in a n c e a n d th e G re a t Recession 691

to which local governments sought credit ratings on TIF bond sales, which accelerated after the financial crisis. In the years immediately before the Great Recession, about 50 percent o f TIF transactions were rated. In the years immediately after the Great Recession, almost 75 percent o f TIF bond transactions received credit ratings. This upward trend is statistically supported by highly significant Pearson’s Chi-squared and log-likelihood Chi-squared statistics o f 62.2 and 59.0. This is likely a combina­ tion o f local governments seeking additional certification o f the value o f their TIF bonds from a third party financial intermediary in the context of the greater financial uncertainty during these years and low credit quality issuers staying out o f the market altogether.

One o f the major calamities in the municipal market during the Great Recession was a collapse o f the bond insurance industry. O f the nine active bond insurance firms before the crisis, only two survived the crisis (Moldogaziev, 2013; Johnson, Luby, and Moldogaziev, 2014). Prior to the financial crisis, almost 60 percent o f all new long-term municipal securities were insured by monoline bond insurers such as FSA, AMBAC, and MBIA (Moldogaziev, 2013). As reported in Tables 4A (2000-2007) and 4B (2008-2013), we find that there is a general increase in the use o f insurance from 2000-2007 with 54 percent o f all TIF bonds insured in 2007. However, there was dra­ matic decline in insured TIF bonds starting in 2008 (only 28 percent insured), which continued through 2012. In 2013, the share o f insured TIFs increased to 34 percent; nevertheless, it is unlikely that bond insurance penetration in the market will achieve pre-crisis levels. Thus, while we would have expected to see an increase in insured TIF bonds after the Great Recession as a means of mitigating investor concerns about bond default, such credit enhancement is not as widely and cheaply available in the capital market as it used to be prior to the Great Recession. As the relationship between the frequency o f insured TIF securities and the year o f issuance suggests (Tables 4A and 4B), there has been a statistically significant shift in the distribution o f insured securi­ ties in the market; the Pearson’s Chi-squared and log-likelihood Chi-squared statistics are 317.9 and 381.0 respectively.

Another dramatic shift in the municipal market was the collapse of the variable rate bond market (Luby, 2012). State and local governments generally have the ability to sell their debt on a fixed interest rate basis where the interest costs are set at issuance or on a variable rate basis where interest rates fluctuate over the term o f the issue. With variable rate debt, state and local governments generally rely on credit enhancement devices such as bank letters o f credit and liquidity facilities to successfully market the bonds over the life o f the issue. Such credit enhancement became very scarce during and immediately after the financial crisis, which led to a significant upward spike in interest costs on these variable rate securities. This phenomenon is shown in our data in Tables 4 A and 4B. It is clear that the volume o f variable rate TIF securities shrunk dramatically after the Great Recession. More than 70 percent o f all variable rate TIFs, or $1.9 billion in our sample, were issued in 2000-2007. This shift appears to be statisti­ cally significant, with Pearson’s Chi-squared and log-likelihood Chi-squared statistics of 27.7 and 33.4, respectively.

692 N ational Tax Journal

T a x I n c r e m e n t D e b t F in a n c e a n d t h e G r e a t R e c e s s io n 693

694 N atio na l Tax Journal

As mentioned above, variable rate bonds and short-term securities, in addition to bond insurance, would also often be accompanied by letters o f credit and/or liquidity facilities. Though fluctuating before and after the Great Recession, the volume o f TIFs using these two forms o f credit enhancements fell after the crisis. As shown in Tables 4A and 4B, the largest volume o f letter o f credit and/or liquidity facility supported securities was issued in 2005, but they have been largely under-used in recent years. In this case as well, the shift in letter of credit and/or liquidity letter enhanced securities is statistically significant. Our estimated Pearson’s Chi-squared statistic is 25.4, while the log-likelihood Chi-squared statistic is 25.6.

IV. D IS C U S S IO N

The Great Recession clearly affected TIF debt finance activity by U.S. local govern­ ments. The volume and number o f TIF municipal securities declined dramatically after 2007. This decline was seen across all states in both the number o f annual transactions and bond volume. Annual TIF bond issuance is now less than half its size (as measured by annual number o f transactions and volume) since its peak in the mid to late 2000s. In addition, the average transaction size o f TIF bond issues has decreased since the Great Recession. This decline in TIF debt activity is most likely the result o f California, the largest issuer o f TIF bonds, slowing down its issuance activity after the recession and ultimately exiting the market, but may also be a result o f other issuers using less TIF debt finance to fund their redevelopment projects given the actual or projected decline in growth o f property tax increments in many parts o f the country.

It appears the type o f projects that TIF debt used to finance has also changed since the Great Recession. Economic development TIF projects rather than general-purpose TIF projects now constitute the great majority o f TIF debt issued since the Great Recession. In addition, the post-Great Recession period has witnessed a total decline in corporate-sponsored private-activity TIF bonds. Our findings provide evidence that bond investors may be demanding more specificity in their investments. There is also evidence o f a lack o f available private capital for government development projects.

The TIF debt market experienced significant changes after the Great Recession with respect to the pricing, structure and sale process o f these securities. Coupon interest rates increased and call maturities declined, reflective o f a general increase in the perceived riskiness o f TIF securities. Local governments increased their use o f general obligation bond structures and independent financial advisors while more often seeking credit rat­ ings on their TIF transactions as a way o f mitigating the perceived increased riskiness o f their municipal securities. Such actions to reduce information asymmetries were especially necessary as other risk mitigation tactics, such as seeking third-party credit enhancement, were not available to these local governments as the monoline municipal bond insurance market collapsed as a result o f the financial crisis.

Given our findings regarding the impact o f the Great Recession impact on TIF debt finance, can we speculate how this financial market will look in the future? Clearly, local governments will continue to sell TIF securities as shown by their continued issu-

T a x I n c r e m e n t D e b t F in a n c e a n d t h e G r e a t R e c e s s io n 695

ance even during the recent global financial crisis. The market is likely to be smaller compared to its historical high since California local governments will be prohibited from selling TIF securities except for refinancing purposes. For example, the year 2012 witnessed an even more dramatic drop than the years immediately after the 2007-2008 financial crisis, as this was the first full year o f California’s TIF debt prohibition. The overall TIF municipal securities market may also be smaller if the cost o f borrowing using TIF debt finance stays higher than other modes o f financing since such additional interest costs will make redevelopment projects less feasible.

It is also likely that the structure and types o f projects financed by TIF debt will change. Local governments appeared to attempt to strengthen the general credit char­ acteristics of their TIF securities after the financial crisis. Such attempts will continue and could take the form o f additional revenue pledges for bond repayment other than the tax increment, general obligation pledges for repayment in addition to the increment, greater use and size o f debt service reserve funds, and higher debt service coverage ratios (i.e., the minimum ratio o f expected tax increment revenues to debt service). Various market strategies related to financial intermediaries may also flourish, includ­ ing the use o f financial advisors and more reputable underwriters and the procurement o f multiple credit ratings, to enhance the perceived credit o f TIF debt. From the bond investor perspective, local governments probably should expect greater due diligence of the credit characteristics o f their TIF securities, demand for more “seasoned” TIF districts, and greater scrutiny of feasibility analyses related to the TIF district. All o f these strategies and demands emanate from the desire o f the investor community for better and more carefully crafted TIF bond credits in an era o f greater risk aversion with respect to the general credit o f local governments.

V. CONCLUSION

Local governments have increasingly relied on the use o f debt finance to raise upfront redevelopment resources in their TIF districts. While the Great Recession certainly “changed the face” o f TIF debt finance, other factors such as the collapse o f the monoline bond insurance industry and the dissolution o f TIF districts in California contributed to the present state o f TIF debt finance. With local governments in Cali­ fornia mostly exiting the TIF municipal securities market and investors likely taking a cautious view o f the issuers that remain in market, the future o f TIF debt finance in the United States is uncertain. However, as long as many local government officials con­ tinue to believe that TIF is one o f their only economic development tools, the financial market should remain active, albeit taking on a significantly different size and market structure.

DISCLOSURES

The authors have no financial arrangements that might give rise to conflicts of interest with respect to the research reported in this paper.

696 National Tax Journal

REFERENCES

Boyd, Terry, 2013. “Insufficient Funds: Arena Debt Document Reveals Struggle to Pay KFC Yum! Center Bonds.” INSIDER Louisville, July 24, http://insiderlouisville.com/news/arena-story/.

Briffault, Richard, 2010. “The Most Popular Tool: Tax Increment Financing and the Political Economy o f Local Government.” The University o f Chicago Law Review 11 (1), 65-95.

Geheb, Philip J.F., 2009. “Tax Increment Financing Bonds as ‘D ebt’ Under State Constitutional Debt Limitations.” The Urban Lawyer 41 (4), 725-754.

Johnson, Craig L., 1994. “An Empirical Investigation o f the Pricing o f Financial Advisor Ser­ vices.” Municipal Finance Journal 15 (3), 36-52.

Johnson, Craig L., 1999. “Tax Increment Debt Finance: An Analysis o f the Mainstreaming o f a Fringe Sector.” Public Budgeting & Finance 19(1), 47-67.

Johnson, Craig L, Martin J. Luby, and Tima M. Moldogaziev, 2014. State and Local Financial Instruments: Policy Changes and Management. Edward Elgar Publishing, Northampton, MA.

Kriz, Kenneth A., 2003. “Comparative Costs o f Negotiated versus Competitive Bond Sales: New Evidence from State General Obligation Bonds.” Quarterly Review o f Economics and Finance 4 3 (2), 191-211.

Lefcoe, George, and Charles W. Swenson, 2014. “Redevelopment in California: The Demise o f TIF-Funded Redevelopment in California and Its Aftermath.” National Tax Journal 67 (3), 719-744.

Leonard, Paul, 1994. “Negotiated Versus Competitive Bond Sales: A Review o f the Literature.” Municipal Finance Journal 15 (2), 12-36.

Luby, Martin J., 2012. “The Use o f Financial Derivatives in State and Local Government Bond Refinancings: Playing with Fire or Prudent Debt Management?” Journal o f Public Budgeting, Accounting & Financial Management 24 (1), 1-31.

Moldogaziev, Tima, 2013. “The Collapse o f the Municipal Bond Insurance Market: How Did We Get Here and Is There Life for the Monoline Industry beyond the Great Recession?” Journal o f Public Budgeting, Accounting, & Financial M a n a g e m e n ts (1), 199-233.

Peng, Jun, and Peter Brucato, Jr., 2003. “Another Look at the Effect o f Method o f Sale on the Interest Costs in the Municipal Bond Market - A Certification Model.” Public Budgeting & Finance 23 (1), 73-95.

Vijayakumar, Jayaraman, and Kenneth N. Daniels, 2006. “The Role and Impact o f Financial Advi­ sors in the Market for Municipal Bonds.” Journal o f Financial Services Research 30 (1), 43-68.

Copyright of National Tax Journal is the property of National Tax Association and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.