This morning Trans Union, the big credit bureau, released its quarterly report on mortgage delinquencies, and it was not pretty. Nationwide, 6.25% of all residential mortgages were at least 60 days past due in the third quarter, up from 5.81% in the second quarter and 3.96% a year ago. This was the 11th straight quarter that delinquencies increased.

Mortgage delinquencies are the first step in a house eventually going into foreclosure, so look for those to start heading up again. Foreclosures have been held down by trial modifications under the HEMP program, but very few of those have gotten to the stage of being final modifications. And even when mortgages are modified, there is a strong tendency for those people to again find themselves in financial trouble. Clearly people not paying on their mortgages is not good news for the big banks like Bank of America (BAC) and Wells Fargo (WFC) that lent them the money.

If there is a silver lining in the data, it is that the rate of increase seems to be slowing. The third quarter increase was “only” 7.6%, which is down from an 11.3% increase in the second quarter and a 14.0% increase in the first quarter. Of course, as the base increases, each additional percentage of increase means a bigger absolute number of delinquent mortgages.

There are huge regional disparities in the rate of mortgage delinquencies. The former bubble states continue to suffer mind-bogglingly high rates of delinquencies — 14.5% of all mortgages in Nevada and 13.3% of all homeowners in Florida are at least two months behind on their mortgages. That is almost one in seven in Nevada and about two in every fifteen in Florida.

In contrast, states where very few people live are experiencing very low rates of delinquencies. North Dakota is holding up best, as it is on a number of economic indicators with a rate of only 1.7%. South Dakota is not faring all that much worse at 2.3% and in Vermont the rate is only 2.6%.

However, the gap is starting to close, and not in a good way. The fastest growth in delinquencies is now coming from areas where there was no real housing bubble. The biggest jump came in Wyoming where delinquencies jumped by 17.9% in the quarter, followed by Kansas at 17.4% and North Dakota at 16.0%. Still, it would take a long time for North Dakota to catch up to Nevada.

There are two key forces that are leading to people not paying their mortgages. One goes to a lack of desire to do so, and the other goes to lack of ability to do so. If your house is substantially underwater, i.e. your mortgage is for a lot more than the house is worth, it is not economically rational to continue to pay your mortgage. After all, most mortgages are non-recourse, which means that the worst thing that happens is that the house gets foreclosed on and you go rent.

At one point, there was a huge social stigma to being foreclosed upon, but as it becomes more common, the stigma diminishes. There are, of course, some non-economic costs associated with not paying and just living rent- or mortgage-payment-free for awhile, and in many areas of the country that can now be well over a year. Your kids might be upset with you since they would have to change schools and leave all their friends if you can’t rent in the same school district. People also develop emotional attachments to their houses. Those factors might be worth a $5,000 or $10,000. However, if the house is underwater by $100,000, most people will just tell little Billy that he will make new friends at his new school.

The second reason for rising delinquencies is unemployment. Quite simply, with no paycheck, it is harder to write the mortgage check. It is not a coincidence that states like Nevada, Florida and California, which have very high delinquency rates, also rank near the top in terms of unemployment — and the Dakotas and Vermont have unemployment rates that are well below the national average. For the delinquency rate to start to fall significantly, we will need to see progress on both the employment front and on the housing price front.

The government has been doing everything in its power to re-inflate the value of houses. It is throwing money at homebuyers in the form of tax credits. Under the recent extension, you don’t even have to be a first-time home buyer to benefit from Uncle Sam’s generosity. Of course, giving money away to move up buyers does not even reduce the inventory of unsold homes, since for each one bought, another one goes on the market.

The Fed has been artificially depressing mortgage rates by buying up $1.25 trillion of mortgage-backed securities. In the absence of that program, rates on a 30-year fixed rate mortgage would probably be at least a full percentage point higher. The Federal government has also assumed the role of subprime mortgage lender through the FHA, which is offering mortgage loans with only 3.5% down, and the tax credit can be used for the down payment. That is exactly the same sort of behavior that got New Century Financial and Washington Mutual into trouble. It just goes to prove the power of a good lobby over economic rationality.

This gift to the realtors of the country is eventually going to come back and bite the country on the behind, resulting in a massive — think Fannie Mae (FNM)- and Freddie Mac (FRE)-sized  bailout — of the FHA.

The massive actions have had some effect, and the Case Schiller index has shown some improvement in housing prices over the last few months. Also, housing prices are much closer to normal, relative to incomes and rents, than they were a few years ago at the peak of the bubble.

Notice that I said “closer to normal,” not “below normal.” In the absence of this extraordinary government support, there is still room for housing prices to fall without them becoming undervalued based on historical relationships. The fact that incomes are not growing much due to high unemployment, and rents are falling due to record high vacancy rates, does not help the situation.

This poses a bit of a dilemma, since new housing starts typically lead changes in unemployment. This can be seen in the first graph below (from In the graph, the unemployment rate is inverted to better show the relationship between it and the rate of housing starts, as well as the lag involved. The crash-induced recession of 2001 is the one case where the relationship does not seem to hold.

The good news is that it looks like we have seen the bottom for housing starts this cycle back in January. Based on the historical relationship, that means we might start to see some improvement in the unemployment rate by this coming spring.

The bad news, though, is that new housing right now is a classic case of mal-investment. With lots of vacant housing, the last thing we need as a country is more housing units. The second graph below (also from shows that the dramatic decline in housing starts has not yet begun to make a dent in the number of houses and apartments that are sitting vacant. Thus it seems unlikely that we will see housing starts return to anything like the 1.1 million a year that has historically been about normal for the country.

More likely the rebound will stall out around the levels that marked the bottom for new housing starts in past cycles of around 600,000 a year. Yes, that is a nice percentage gain from the lows of under a 400,000 rate, but it does not suggest a robust recovery.

Read the full analyst report on “BAC”
Read the full analyst report on “WFC”
Read the full analyst report on “FNM”
Read the full analyst report on “FRE”
Zacks Investment Research