Demographics and Jobs

Take a look at this interesting picture sent in by NorthGG. It shows the month on month gain in non-farm payroll (NFP) against the twelve-month growth in the workforce. I know workforce growth is somewhat endogenous, but nevertheless, the relationship is good with workforce growth lagging just a bit behind jobs.

I take away two lessons from this picture. First, and by far the most important, demographics matter. In the 1970s, the first cohort of baby boomers is just entering the workforce en masse. As a result, independent of the point in the cycle, workforce growth is relatively high. This means that equilibrium job growth also has to be high. By the 2000s, the first cohort of boomers is starting to retire. Equilibrium job growth is low. Losing 500,000 jobs in 1970 is different than losing 500,000 jobs now. Since fewer new entrants need jobs, we need fewer created jobs each month. 500,000 lost jobs creates a smaller "jobs gap" now than it did in 1970. Its still bad, just not as bad as it used to be.

Second, workforce growth lags the cycle a bit. We should not expect to see the bottom in terms of job losses until the twelve-month growth rate of the workforce starts to plummet to zero. I see a little bit of downward movement in the picture above but not a nosedive. We are not yet at the bottom.

Thanks again to NorthGG for sending in the intriguing picture. I will point you toward three interesting economic papers on demographics. First, a recent paper in ReStat studies the relationship between productivity and demographics. Second, Mankiw and Weill explained the 1990s downturn in house prices using demographics. This issue is explored further in a Fed paper from a few years ago. I like the way this paper compares the situation in the United States to that in Japan.

I believe the effects of demographics, in particular the effects of exogenous fertility shocks on the labor force, are among the most understudied factors in economics. Of course, many economists don't believe fertility shocks are exogenous. They are probably right.

Hope for the Housing Market

And, if there is hope for the housing market there is hope for the economy.

Almost three years into the housing downturn (remember Housing Starts first turned negative in January 2006), the inventory of new homes for sale is finally starting to adjust in a meaningful manner.

Take a look at the picture below. As is typical during any stock adjustment, inventories rose during the first few months of the downturn. What is not typical is the behavior of the stock of unsold homes over the next 18 months. Inventories of homes for sale did not decline whatsoever, on average, between January 2006 and the summer of 2007.



Of course, as we have learned with so many series during this recession, falling to a low level does not mean the series is near the bottom. I don't have a crystal ball, but we can look at the series during previous downturns and make judgements about where it is likely to end up in this downturn.

Here is a long time series of houses for sale.

Housing inventories have a long way to go before they finish their adjustment, but the end may be in sight. If the number of homes for sale continues to fall at its 2008 average pace, the inventory of unsold new homes will hit its bottom (maybe 250k?) in March 2010 or in about 16 more months.

I am assuming that inventories only need to fall as far as they did during the 1982 downturn. As of right now, this recession looks worse than that one so it may have to fall farther but this number feels right.

The number of months left seems very large. That is why inventory overhangs are such a problem--remember the fiber optics networks in 2000. But, even if I use a much larger number for adjustment, we still need several months to work through the excess. For example, if I use October's monumental fall of 8 percent, the inventory adjustment takes another 6 months. No way am I that optimistic.

There is a silver lining, however, even in the 16-month adjustment scenario, my base case.

Housing starts stop falling long before the inventory adjustment is complete (see chart below - starts are the jagged black line). In every previous housing-market downturn, housing starts stopped falling about one full year before the inventory of new homes reached its bottom.

And, once housing starts stop falling so does the labor market. This means in about three months (four more labor reports), sometime in the very early spring, the labor market will stop getting worse: we will reach the worst of the month-on-month declines. A few months after that employment will start growing again. According to this forecast, we will see positive job growth as soon as next summer.

Does this mean the recession is going to end in mid-2009? Maybe.

Maybe not. All of the above is predicated on this downturn looking something similar to all of the other post-war recessions. The gorilla in the room right now is the large number of foreclosures. These foreclosures are adding to housing inventory in a manner very similar to new housing (they are empty and must be sold) and the number of foreclosures is at record levels. Therefore, these estimates may be wildly optimistic. My guess, and it is nothing more than a guess, is that the foreclosure crisis is good for about 4 to 6 months extra adjustment.

We shall see.