Thursday, September 20, 2007
Freight And Employment
It's important to remember the distinction between overall unemployment and insured unemployment, which is less than half of overall unemployment. The relationship varies with changes in the economy (for example, if unemployed workers fall off the rolls, they will no longer show in this report, although they should be reflected in the household survey), casual employment, contract workers, etc. Most construction unemployment is by contract and therefore isn't reflected all that well in these figures. Also, if a worker gets severance, their unemployment coverage doesn't begin until after that runs out, so these numbers are not necessarily as topical as the monthly employment reports.
The mid August through mid September unemployment claims, especially continued claims, are always heavily affected by the elementary, high school and colleges moving back into full swing. Last year from 8/12 to 9/16 (you can get these figures here, choose national and the year ranges you would like), initial SA claims were reported in a very tight range extremely similar to this years', but NSA continuing claims fell from 2,326,364 to 2,062,962, with continuing claims dropping about 90,000 between 9/9 and 9/16.
As for freight, August rail figures are still discouraging:
U.S. railroads originated 1,685,238 carloads of freight in August 2007, down 17,008 carloads (1.0 percent) from August 2006. U.S. railroads also originated 1,195,390 intermodal trailers and containers in August 2007, a decrease of 52,263 units (4.2 percent) from August 2006, the Association of American Railroads (AAR) reported today.Compare that to the YTD:
For the first eight months of 2007, total U.S. rail carloads were down 414,977 carloads (3.5 percent) to 11,367,593 carloads. Year-over-year traffic is down in most commodity categories, including crushed stone, sand, and gravel (down 85,941 carloads, or 10.6 percent); coal (down 76,095 carloads, or 1.6 percent); and motor vehicles and equipment (down 50,384 carloads, or 6.7 percent).Trucking figures come out one month behind rail but represent a much larger percentage of shipping. The latest we have is July:
U.S. intermodal traffic, which consists of trailers and containers on flat cars and is not included in carload figures, was down 154,217 trailers and containers (1.9 percent) for the first eight months of 2007 to 8,061,355 units.
This is not an encouraging graph. On one hand, the recent trend line is better than it was for Sept-Jan of last year, but on the other hand, it is below 06, 05 and most of 04. Volume does matter; most economic measures are reported in currency, but if you look at the graph above you will see that the volume is closely associated with jobs.
Other disturbing trends in recent rail and trucking freight stats are the intermodal and the number of for-hire truckloads, which roughly correspond. Intermodal rail captures a lot of import to store measures, and recently have showed more weakness than in the beginning of the year. For-hire truckloads actually showed a slight increase in July, and that combined with lower tonnage tells us that the point-to-point shipments are dropping in volume. This is going to be highly related to lower actual retail sales and possibly lower manufacturing, although we do not know that yet (however, an Industrial Production report that would have been negative except for utilities and high consumer electricity demand is a hint).
Currency-denominated economic measures can easily miss the initial stages of economic downturns in inflationary circumstances; employment measures, as I have discussed before, are very affected by the circumstances in the prior year, and therefore also often miss the early stages of economic downturns and upturns. CR at Calculated Risk is a very good economist and he has taken great pains to post predictive data and also discuss the inherent uncertainties in these measures, which most people do not grasp.
The reality is that we don't know actual employment for at least a year, and we don't have accurate inflation adjustments on most of our economic reports. Employment is a lagging to coincident indicator anyway, and therefore not predictive. The way household employment is calculated can and does miss substantial changes. Consider the difference, for example, between a construction worker employed on average 2 days a week as opposed to 5 days a week. As far as the household survey is concerned (even after B/D adjustments), that is one job. As far as the effect on the overall economy is concerned, that is one job paying a net of about 45% less (you have to adjust for lower payroll and income taxes).
What is predictive are volume measures like those I am showing you here, and relationship measures, such as those CR has documented for the relationship between equipment and software and fixed investment.
There are aspects of the economy that do not show up in freight measures. Those are jobs like information. But one thing we can confidently predict is that the information sector will be badly hit by the problems in finance and mortgages; a huge part of the boom in information was related to all of those systems for all of those companies, quite a few of which are now out of business. Another predictive aspect is that too much job growth has been deriving from government and education, which are derivations, over time, of growth in private income and jobs. It is the construction worker, the retail clerk, and the trucker who pay the salaries of the teachers and the police, not vice versa. For a number of reasons, we are safe in forecasting that this trend cannot continue for very long, and indeed in August a notable weakness in government employment appeared.
The economic picture one sees if one includes volume and relationship measures is very different than if one just looks at the headline reports. For this reason, I expect overall inflation to be less of a factor than many think it will. The reality is that the economic basis for passing through costs to consumers is weak, and all history tells us that when you see these pressures, inflation is coming under constraint. It would take huge stimulative measures to push it out of control, which have not happened yet. Another factor people need to grasp is that a credit contraction is a deflationary force of no mean proportions. Debt expands the effective money supply; a contraction in debt reduces it.
So inflation (especially inflation on imports related to weaker dollar) is real factor, but it is not the only factor. It's a sad fact that almost all economic reporting to the public concentrates on the least certain and least determinative measures of future performance. It's no wonder that everyone always gets caught by surprise.
Update: Oh, yes, commercial paper outstanding is still dropping. You can see this post for the updated graph (which is a direct link). Why Bloomberg feels this is a surprise is completely beyond my grasp. Total is now 354 billion, and the weekly was 48.7 billion. AA financial is still increasing, and financial's continued decline is a function of crappy fundamentals and higher prices. In short, people's appetite for crappy debt at low yields has not been affected at all by the rate cuts. Why it would, especially in a day and a half's time, is beyond my comprehension. Next week's numbers matter more.
I won't take any credit but I've been harping on ton-miles over on CR for a long time. You need to do a little adjustment for things like coal (weather) and lumber (housing sector) but man what a great general ecnomic signal. Look at a 3mo MA and see the future.
I think you are correct about inflation, but here's the puzzle:
1) Some components of inflation are purely international, and thus not subject to vagaries of US monetary policy;
2) Some components are affected by a weakening dollar, but a weak dollar is needed to gain in exports, which is needed over time to rebalance the economy and support the dollar;
3) The effects of a credit contraction on top of what looks like genuine developing weakness are likely to be anti-inflationary in the domestic economy.
I think some people presume that there is a right and perfect answer, but my belief is that there isn't. We are stuck with the fruits of our own bad economic policies, and the best the Fed can do is probably try to steer between equal and opposite dangers. That is the way I read their last statement.
The inflation from consumption; food and energy will eventually filter back to the US economy.
There is no current trend crossing of imports and exports that restores nevermind reverses the imbalance of trade.
consumption and ultimately will reduce
Rob, but eventually there will have to be. You've written a lot about some of the misled stuff about mass transit and other malfeasances of public policy. I have enjoyed your posts a great deal, and learned quite a bit. My feeling is that quite a bit of our current situation is related to energy, and a lot of our current energy situation is related to completely dysfunctional policies which have over time eroded our ability to generate energy. If we had continued on the trajectory of nuclear power and other domestic sources, our current trade imbalance would look dramatically different.
There are number of applications for nuclear power plants, and I think reality will slowly seep through the hysterics and bring about a correction. If not, Rob, we are going to be very poor, and Americans don't like being very poor. Sooner or later, they are going to figure this out.
There is an indirect impact, in that to the extent that nuclear reduces the power-plant demand for nat gas, then gas may become a more economical heating source and displace some oil use in that application.
To get a big oil-use reduction from nuclear, we would need to apply electricity as a transportation fuel, which could mean (1)plug-in hybrids, (2)hydrogen, and/or (3)electric railways. I think (1)is the most practical in the reasonably near term, perhaps with niche applications for (3) as well.
Links to this post: