When it comes to the over and under on unemployment, I'll take the over. As I conveyed in January when I discussed some of the implications of a shrinking global economy, I believe it is a foregone conclusion that the rate of unemployment will climb towards 10% this year, but I am sticking to my suggestion that it actually hits 11-12%. In terms of job loss, this implies that we will lose a total of 5-6mm jobs, or an average of about 500K per month. Given that this process should be front-loaded to a great degree, we run the risk of a record monthly decline of 1mm jobs. The forecast for January is for a decline of 500K. I have no clue if it will be this month or next, but I don't think the market or our collective confidence are prepared for a headline like that.
A few caveats:
- I am not an economist. I have no fancy models, but I have been following the data closely now for 23 years
- The data is subject to revisions and new benchmarking and has substantial statistical margins of error
I see several factors suggesting a watershed event like a 1mm monthly job loss.
- Pent-up Supply of Pink Slips
- The Small Business Effect
- The Cycling of the Numbers
- Historical Precedent
I had a thesis last year that proved correct initially but that ultimately failed as the economy went over the cliff. Looking back at the very tough times companies had finding qualified labor after they drastically cut payrolls following 9/11, I had expected that they would be very reticent to lay off during this downturn. This proved to be the case until the fall. While it's not especially clear in the chart below over the past 50 years, it wasn't until September that we had a monthly decline in excess of 127K. The declines in the first 8 months of the year, well into the official "recession" that has now been dated, were, in fact, stunningly low. Of course, the floodgates opened late in the year, and we have been running at a torrid pace now for what will be the 5th consecutive month. I believe that we are seeing a proverbial throwing in of the towel by large companies who are in the process of evaluating the new economic environment and abandoning their desire to carry their workforce through the downturn.
When we read the papers or listen to the news, it is quite evident that large companies are laying off workers en masse. A more subtle trend to detect is what the even larger overall employment providers, small and medium sized businesses, are doing. I have encountered several anecdotal signs of a rapid retrenchment, and it makes sense. As smaller companies see pressure on sales and margins, the owner is able to manage costs by working harder. Additionally, spouses, family or friends who become unemployed or underemployed can help pick up the slack as the owner cuts the payroll. Shrinking credit availability isn't helping make it any easier for these smaller businesses to hang onto employees.
As I mentioned above, I expect to see a total job loss of about 6mm this year, which equates to 500k per month or a continuation of the trend of the past 3 months as well as the expectation for this month. I assume that a lot of this will be front-loaded and expect that at least half of the losses will occur in the first 4 months of the year, suggesting that the average will be 750k per month through April. If the economists forecasting "just" 500k hit it right this month, that suggests the average over the next three months would be about 850K. Clearly, we could hit 1mm in one of those months. While we are hearing of the massive layoff announcements, it is hard to predict when they actually hit. It would seem that the biggest headline risk would be in February.
My final basis for being open-minded regarding the potential for a month of 1mm jobs lost is looking at it as a percentage of the base rather than an absolute number. In the chart above, we had an increase of 1.1mm jobs in September, 1983. The largest decreases over the past 50 years have been 602K in December, 1974 and the recent 584K in November. If, however, one looks at the data as a percentage of the base of employed non-farm workers, it is clear that there is precedent for a higher number. There have been two instances of gains in excess of 1%, while there have been several instances of losses of .5% or more. If one takes the current base of 135mm and assumes a 0.8% decline (like the worst print in 1974), we would see a loss of almost 1.1mm jobs.
The rapidly deteriorating employment situation is so powerful that it will be creating a negative feedback loop that will adversely impact the overall economy. Many of our safety net programs are funded via payroll taxes (Social Security, Medicare) or through corporate payments to state governments (unemployment insurance). Income taxes will also be impacted. At the same time that government income is threatened, demand for services will be escalating. As you can see in the chart below, the duration of unemployment has been increasingly longer and for more extended periods for quite some time, and we are risking leaving a large segment of our population as permanently unemployable. The statistics cover up another ugly truth: Many folks are underemployed. We are beginning to see downward pressure on wages as well, and clearly the workweek is shrinking. Clearly, consumption will be pressured for quite some time as unemployment rises and incomes stagnate due to either wages falling or fewer hours worked.
The government can do very little to change the trends now firmly in place. In the long run, it is crucial that we retrain those workers who no longer are able to work using the skills that they have acquired, but this takes a long time. While it makes little sense to keep people on the job when the job isn't economically justifiable, it seems like many of the job losses in the coming year will be from viable companies just scaling back for an extremely weak economy. It may make sense for the government to promote job retention by subsidizing it. An easy way to do this would be to eliminate employer contributions to SS and Medicare, which would effectively lower the marginal cost of keeping employees. Similarly, the government might consider subsidizing unemployment insurance and worker's compensation insurance that is paid by employers to the states. These programs would effectively pay for themselves to the extent that they reduced the loss of payroll (employee contribution) and income taxes and served to improve corporate cashflows.
While I could be wrong about actually getting a print of 1mm jobs lost in a single month, I think that the odds are pretty strong. This economy is the worst since the 30s, and the data indicates that the monthly employment change can drop by as much as 0.8% in a given single month. The more important issue is to understand that the deferral of layoffs last year will result in a continued near-term assault on the workforce. The economic implications are extremely negative. Until the market begins to better understand that unemployment will move into the double-digits, stocks are most likely not properly discounting the risks to even lower earnings.