Do CDFIs Hit Their Targets?

by Oscar Gonzales and James Greer

The authors recently published an important monograph on CDFIs, the first analysis of its kind. This is the second part of their summary of key findings.

The authors wish to stress that the opinions and views expressed here are theirs and theirs alone and do not reflect the position of the Department of the Treasury or the Federal government.


In our first contribution to moneypluspurpose.com, we summarized the multiple sources of market failures that define the difficult environment in which CDFIs and CDEs (Community Development Entities in the New Markets Tax Credit, or NMTC, program) operate, a subject that we examine in the first chapter of our monograph.

In central city low-income (and frequently minority) communities as well as in many rural areas, CDFIs provide financial services and makes many loans and investments in communities that have been for decades in the backwaters of economic and social development across this country.  CDFIs foster economic and housing development in some of the most challenging areas of the U.S., communities ravaged by disinvestment, the systemic under-provision of public goods, especially the basic infrastructure required to initiate and sustain economic growth and housing development. CDFI markets generally are areas that have suffered for generations from poverty. They are typically places where, if only because mainstream financial institutions have made precious few loans and investments, there is little or opaque credit histories of individuals and businesses upon which to make reasonable investment decisions by the nation’s banks or other mainstream financial institutions.

While we only provide an abbreviated version here of a more systematic and detailed investigation of the portfolios of the many CDFIs and CDEs, the main outlines of our findings provided in chapters 4 and 5 of our monograph can be quickly summarized. Bluntly, these findings are quite striking:  the CDFI industry has, over the past several years, successfully made extensive investments in some of the most distressed communities across America: very low and low-income areas, impoverished communities, and places that are largely occupied by minorities, including Native communities. Indeed, compared to the investments of mainstream financial institutions, CDFIs and CDEs have focused their investments in American neighborhoods into, what by any standard, are highly distressed communities across the country.

Bluntly, these findings are quite striking:  the CDFI industry has, over the past several years, successfully made extensive investments in some of the most distressed communities across America

This first table (Table 1, below) provides a powerful summary of both the magnitude and differences between the investments made by mainstream financial institutions in contrast to the CDFI and NMTC industries.  Here we aggregate investments by the median family income of census tracts providing the total population (in 2000) of tracts by income and then sequentially the total dollars in home mortgages, home improvements, and refinancing by mainstream financial institutions using the Home Mortgage Disclosure Act (HMDA) data. We compare that to transaction information on CDFIs and CDEs (under the New Markets Tax Credit program) derived from data collected and annually published by the CDFI Fund as part of its Community Investment Impact System, or CIIS .  The contrasts presented in this table are striking.

The most significant finding is that CDFIs and CDEs have invested primarily in very low and low-income tracts in both urban and rural areas. In stark contrast, mainstream financial institutions have consistently eschewed low-income areas and have concentrated their investments disproportionately in middle and upper-income communities across the country.

The most significant finding is that CDFIs and CDEs have invested primarily in very low and low income tracts in both urban and rural areas

Table 1: Mainstream and Community Development Lending by Median Family of Tract, Rural and Urban Areas
Population by Median Family Income of Tract Population HMDA Originations ($1,000s) CDFI Transactions NMTC Transactions
Very Low/Low 92,595,859 1,829,667,591 3,420,625,321 16,762,852,148
Moderate 131,840,799 5,049,744,348 1,577,019,874 2,255,895,541
Middle 51,446,772 3,973,063,191 490,442,955 741,825,620
High 5,008,032 662,613,452 97,448,836 78,490,000
Total 280,891,462 11,515,088,582 5,585,536,986 19,839,063,309
Very Low/Low 33.0 15.9 61.2 84.5
Moderate 46.9 43.9 28.2 11.4
Middle 18.3 34.5 8.8 3.7
High 1.8 5.8 1.7 0.4
Source: Greer and Gonzales, Community Development in the US, Table 5.2

The contrasts are marked: whereas banks, mortgage companies, and other mainstream financial institutions have invested just 16% in very low and low-income communities, CDFIs have in contrast made more than 60% of their overall investments in these communities, and CDEs under the NMTC program made nearly 85% of their investments in very low and low-income neighborhoods. Middle-income communities received more than one-third (34.5%) of HMDA investments while such areas were the site of just 8.8% of CDFI loans and a very small proportion of NMTC investments (3.7%).

The data presented in Table 1 also highlights another fact: the total amount of loans and investments made by mainstream financial institutions dwarfs that of the community development industry.  Namely, the total dollar amount of home mortgages, home improvement, and refinancing loans over the five-year period reported here sums to more than $11.5 trillion dollars while CDFI investments total $5.6 billion and the NMTC program nearly $20 billion.  While $25 billion in community development is clearly a substantial amount, it pales in comparison to the totality of investments by mainstream financial institutions. These CDFI and CDE investments can provide only the most modest (indeed very modest) engine of economic and housing development to the nation’s economically distressed communities in comparison to the massive infusions of investment routinely made (albeit unevenly especially after the onset of the Great Recession in 2008) by the nation’s banks and other mainstream financial institutions.

Policies that could even modestly divert the portfolios of banks and other regulated financial institutions to the needs and prerequisites of development in the nation’s economically distressed regions would likely produce the basis of extensive growth in those communities.

All CDFIs have consistently focused their investment activities primarily in very low (less than 50% of the median metropolitan or state family income) and low-income areas (50% to 80% by the same criteria).

One question that readers of our monograph might reasonably ask is if there are any noticeable differences between regulated (banks and credit union) CDFI investments in comparison to the numerous unregulated loan funds and venture capital firms? Overall, we show that CDFIs, in fact, make the predominant share of their loans and investments in distressed communities, but is this consistent across the CDFI industry?  The answer is clear in Table 2, where we provide a summary of the percentage of total dollar investments by financial institution type over five of the past years.   From the evidence in this table, all CDFIs have consistently focused their investment activities primarily in very low (less than 50% of the median metropolitan or state family income) and low-income areas (50% to 80% by the same criteria).  A much smaller portion of bank, credit union, loan and venture fund dollars (and generally a declining share over this five-year period) are originated in middle and upper-income census tracts.

Table 2: CDFI Investments, by Financial Institution Type, in Very Low & Low and Middle & High Income Areas
2010 2011 2012 2013 2014
CDFI Banks
Very Low/Low 32.7 35.2 32.8 26.7 30.1
Middle/High 11.9 16.6 16.4 6.6 9.6
CDFI Credit Unions
Very Low/Low 31.6 32.6 31.8 30.4 30.5
Middle/High 13.1 9.8 11.3 3.2 3.0
CDFI Loan Funds
Very Low/Low 41.3 42.7 44.4 38.7 42.0
Middle/High 10.2 8.3 6.3 3.5 4.0
CDFI Venture Funds
Very Low/Low 46.6 46.0 48.2 40.2 25.7
Middle/High 0.4 1.7 0.2 3.3 5.6
Source: CDFI Fund, CIIS transaction-level report data

Why should any significant share of CDFI investment dollars go to these ostensibly better off communities?  It is important to note that median family income of tracts is only one of several criteria that determines whether any one individual tract is considered, by the CDFI Fund, as an “investment area”; also relevant is the poverty level, the proportion of minorities, or “other targeted populations” and indeed other criteria as well.  Nestled within middle and even higher income tracts there often are small pockets of economic distress that many CDFIs are committed to serve. As a result, a relatively small proportion of CDFI investments flows to such areas. Nonetheless, by far the highest proportion of loans and investments by the many CDFIs that report their portfolios to the CDFI Fund through the CIIS are located in economically distressed communities. Overall, the contrast to lending by mainstream financial institutions is both consistent and striking.

Indeed, as we show in Table 3, housing lending by mainstream financial institutions over the past several years has consistently flowed much more to middle and high-income communities, a marked difference from CDFI lending. Recalling that only about 20% of the nation’s urban and rural populations reside in these middle and high-income tracts, consistently mainstream financial institutions have lent about 40% of their total housing mortgage, home improvement, and refinancing dollars to these areas.

Another interesting fact, all too evident for even the casual observer in the years following the Great Recession, is that the total dollar amount of loans by banks, thrifts, and increasingly mortgage companies, has fallen precipitously:  in 2008, the pinnacle of the crisis and when lending was already declining, the total dollar amount reported in the HMDA data was just under $950 billion; by 2013 that amount had dropped to just over $250 billion—and these trends continue to this day.

Table 3: Home Mortgage Disclosure Data (Mainstream Financial Institutions) Lending by Income of Tract
             
  2008 2009 2010 2011 2012 2013
             
Very Low/Low 163,004,328 134,973,077 66,034,202 36,695,875 34,152,382 40,813,187
Moderate 419,912,587 383,523,268 216,344,253 122,397,786 109,618,593 110,239,421
Middle 189,195,549 177,327,518 108,590,167 65,401,234 56,677,778 50,715,492
High 166,662,547 150,738,970 106,102,491 70,224,259 61,901,165 53,782,874
             
Total 938,775,011 846,562,833 497,071,113 294,719,154 262,349,918 255,550,974
             
             
Very Low/Low 17.4 15.9 13.3 12.5 13.0 16.0
Moderate 44.7 45.3 43.5 41.5 41.8 43.1
Middle 20.2 20.9 21.8 22.2 21.6 19.8
High 17.8 17.8 21.3 23.8 23.6 21.0
             
Source: Federal Financial Institutions Examining Council      
             

The last area we want to highlight—examined at length in Chapter 4 of our monograph—is tax expenditures, a subject you are likely going to hear much more of in the coming four years during the Trump administration.

Tax expenditures or tax credits are as old as the American tax code. Beginning with the creation of an income tax system, homeowners were permitted to deduct the interest on their mortgage payments.  Initially, this was a small portion of the federal budget, but over the past century this and similar incentives have become embedded in the tax code have come to be an ever larger and permanent feature of American fiscal policy. Tax incentives are now employed to reduce the cost of local property taxes, to provide child care, to enhance the production of low-income rental housing, to encourage education, to underwrite the use of alternative and sustainable forms of energy, to provide an income floor for low-income workers, and, as most readers of this blog are very familiar, to enhance the development and expansion of businesses and encourage commercial developments in low-income areas under the New Markets Tax Credit program.

Beginning with the creation of an income tax system, homeowners were permitted to deduct the interest on their mortgage payments.  Initially, this was a small portion of the federal budget, but over the past century this and similar incentives have become embedded in the tax code have come to be an ever larger and permanent feature of American fiscal policy.

Table 4 provides an abbreviated summary of tax expenditures.  Again, we focus on the distribution of these resources by the income of areas, here zip codes (the only aggregation of tax expenditures publically available).

Table 4: Tax Expenditures Estimates, by Income of Zip Code, 2012
Income Level of Zip Code Total Tax Expenditures Mortgage Interest Deduction Local Taxes Deduction Real Estate Taxes Deduction Earned Income Tax Credit New Markets Tax Credit
($1,000s) ($1,000s) ($1,000s) ($1,000s) ($1,000s) ($1,000s)
Very Low/Low 43,671,791 8,658,868 5,487,761 3,141,979 25,274,897 1,108,286
Moderate 27,622,024 8,316,527 4,728,977 3,183,229 11,230,097 163,195
High 50,903,537 18,826,556 10,662,551 7,821,892 13,425,184 167,355
Highest 131,672,448 51,223,723 42,062,193 27,629,804 10,682,891 73,837
         
253,869,800 87,025,673 62,941,482 41,776,904 60,613,069 1,512,672
Percentage
Very Low/Low 17.2 9.9 8.7 7.5 41.7 73.3
Moderate 10.9 9.6 7.5 7.6 18.5 10.8
High 20.1 21.6 16.9 18.7 22.1 11.1
Highest 51.9 58.9 66.8 66.1 17.6 4.9
100.0 100.0 100.0 100.0 100.0 100.0
Sources: http://www.irs.gov/uac/SOI-Tax-Stats-Individual-Income-Tax-Statistics-2012-ZIP-Code-Data-(SOI);  CDFI Fund; Greer and Gonzales, Table 4.2.
 

What is most evident from the data provided in Table 4 is the striking inequality of tax expenditures in the United States, a trend that has persisted for decades.  Overall, based on this one year of information, total tax expenditures are noticeably biased towards the highest income areas of the country: zip codes where incomes are the very highest absorb over half of all the total dollar amount of these credits with another fifth in high-income areas.  This occurs because of the very large components of tax expenditures in the IRS code—the interest deduction for home mortgages, as well as deductions taxpayers make for local property and sub-federal income taxes—all consistently benefit, high and very high-income areas.

In contrast, the New Markets Tax Credit, a program administered by the CDFI Fund, as per the intent of the legislation that created the program (a late Clinton administration initiative) concentrated the use of these tax expenditures to very low and low-income communities in both urban and rural areas across the country.  Indeed, we emphasize, as is all too obvious in Table 4, that NMTC investments are overwhelmingly focused on economically distressed areas of the country.  The New Markets program channels most of the investments leveraged by this program into very low, low, and moderate-income neighborhoods (about 85% of these investments) representing in negative a mirror of the concentration of nearly all tax expenditures in the U.S.

The New Markets Tax Credit, a program administered by the CDFI Fund, concentrated the use of these tax expenditures to very low and low-income communities in both urban and rural areas across the country.

 

 

2 comments

  1. Excellent summary ! We can post this on http://www.ethicalmarkets.com if you like

    Liked by 1 person

  2. mpinskyllc · · Reply

    Please feel free to re-post, Dr. Henderson. Thanks for your interest.

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: