The Diversity Recession: A debunking

UPDATE: 12/31/2009: The paper by Laderman and Reid of the San Francisco Federal Reserve Bank provides the crucial information on foreclosure rates in California, the heart of the mortgage meltdown:

We also find that race has an independent effect on foreclosure even after controlling for borrower income and credit score. In particular, African American borrowers were 3.3 times as likely as white borrowers to be in foreclosure, whereas Latino and Asian borrowers were 2.5 and 1.6 times respectively more likely to be in foreclosure as white borrowers.

So, in the economists’ simple multiple regression model, after adjusting for income and FICO, minorities in California still had substantially higher foreclosure rates than whites:

- blacks 3.3X
- Latinos 2.5X
- Asians 1.6X

(These adjusted gaps are all statistically significant at the 0.01 level.)

Presumably, the raw differences in foreclosure rates are greater. Unfortunately, the actual raw numbers aren't listed in the report, and the authors refused my repeated email requests to release the unadjusted numbers by ethnicity.

The raw ratios are important for estimating the overall share of defaulted dollars by ethnicity in California. We know from the federal HMDA data that minorities accounted for 77% of subprime home purchase dollars borrowed in California in 2006 (the worst vintage for defaults) and 56% of all home purchase dollars. You can see the graphs here. (I'm excluding borrowers of unknown ethnicity and mixed ethnicity couples).

In other words, minorities accounted for the great majority of defaulted dollars in California. And, California accounted for a sizable majority of all the defaulted dollars that launched the mortgage meltdown that launched the Great Recession.

For the story of my request under the Freedom of Information Act to obtain a copy of the unadjusted ratios, see here.

UPDATED September 27, 2008: For a subsequent rebunking of the connections between the Bush Administration's ideology of "diversity" and the mortgage disaster, see my article on "The Man Who Has Escaped All Blame for the Mortgage Meltdown ... So Far."

Since writing that fairly comprehensive article on Saturday, I've also found some seemingly trustworthy statistics on recent subprime loans by ethnicity. (Unfortunately, nobody seems to have calculated defaults by ethnicity. Perhaps nobody wants to know?)

We do know that defaults are closely tied to subprime loans. The most toxic of all, adjustable rate mortgage (ARM) subprime loans, accounted in early 2008 for only six percent of all loans outstanding but 39 percent of foreclosures started. Fixed and adjustable subprimes account for only 12 percent of loans outstanding, but half of current foreclosures. The subprime share of new lending roughly doubled from 2003 to 2004 and increased again in 2005. A remarkable fraction of defaults appear to have come from mortgages originated in 2004 through early 2007, when median prices in California, Nevada, Arizona, and Florida were absurdly high relative to median incomes.

Compliance Tech, a firm that helps lenders "Manage Diverse Lending Markets," estimates that in 2004-2006, minorities accounted for 44 percent of all subprime loans, with Hispanics slightly outnumbering blacks.

Minorities probably have higher default rates than whites. For example, default rates on college loans are about five times higher for blacks, and more than two times higher for Hispanics, than they are for whites. (Asians are best of all at paying back college loans.) Normally, college loans are handed out more willy-nilly than mortgages, so the ethnic default rate gaps likely aren't as big in mortgages, but willy-nilly pretty much describes 2004-2006 mortgage lending in some markets.

On the other hand, they probably tend to have lower value mortgages. On the other other hand, many Hispanics are found in expensive states, especially California, epicenter of the crisis. The largest defaults in America as measured in dollar losses (number of defaults times size) appear to have come out of California's exurban fringe: the Inland Empire, Antelope Valley, and the Central Valley. All these areas tend to have mixed white and Hispanic populations.

Now, that earlier debunking.

A friend emails:

Quite a lot of mainstream conservatives---seemingly following Steve's lead---are starting to argue that racially oriented government housing policies played a very large role in the ongoing collapse of America's financial system, e.g.
The Diversity Recession:
http://www.vdare.com/sailer/080921_cornerstone.htm
and many, many, many other articles.

I just can't see how this makes any sense. Here are my back-of-the-envelope estimates:

(1) Blacks+Latinos together make up about 1/4 of America's adult population.

(2) A very disproportionate fraction of blacks+Latinos just aren't in the mortgage/homeowner demographic category. After all, a substantial fraction of adult blacks are either current convicts, drug addicts, or welfare recipients, and a substantial fraction of adult Latinos are either recently arrived or otherwise improverished illegal immigrants.

While some members of these sub-demographic categories do own homes and have mortgages, the rates are surely extremely low compared with the general population. For example, I really don't think too many of those Latino immigrant day-laborers at Home Depot own their own homes.

(3) Furthermore, even if we exclude those black+Latino subgroups, the remaining black and Latino population is considerably poorer than the
white+Asian population. Poorer people are much more likely to rent (or
live casually with friends and relatives) rather than own a home and have a mortgage.

(4) Therefore, I'd guess that that although black+Latinos are about 1/4 of the adult population, they'd probably have only about 10-15% of the home mortgages. Presumably, there must be some official statistics on this somewhere.

(5) Next, consider that blacks and Latinos tend to live in heavily black and Latino areas, where the housing prices are probably far below the national average. And since even homeowning blacks and Latinos are almost certainly poorer than homeowning whites+Asians, they're homes are cheaper. And probably very, very few blacks or Latinos can afford the upper-end $1M+ homes.

(6) Therefore, my rough guess is that while blacks+Latinos might hold 10-15% of the home mortgages, the total dollar amount of the mortgages held by those groups is much lower, perhaps just 5-8%. Admittedly, many Latinos live in ultra-expensive CA, but very few blacks do, and housing in heavily Latino TX is pretty cheap. A big fraction of the black population lives in the Deep South, where home prices are very low.

(7) Now suppose that 5-8% of the dollar value of American mortgages is held by blacks+Latinos. And let's suppose that the default rate among
blacks+Latinos is 3 times the white+Asian average, which is probably on
the very high side. If so, that still means that the impact of disproportionately high black+Latino default rates on the national dollar amount default rate would be pretty negligible, not remotely the sort of thing that would be causing all those banks (and maybe our entire financial system!) to collapse.

(8) Furthermore, the slice of the black+Latino population with supposedly the much higher default rates, e.g. the subprime borrowers, are also paying much higher interest rates and various extra fees and penalties, which would partially mitigate the impact of the higher default rate.

Anyway, that's my very crude analysis. I'd be very interested in knowing where my numbers might be off or what a better set of estimates might look like.

Any comments?

My published articles are archived at iSteve.com -- Steve Sailer