It's been awhile since we last revisited unemployment. I have largely abandoned its study because of the distortions that have been built into its present calculation. These distortions become clear in looking at the labor force participation rate, which is the proportion of the adult population engaged in full time employment, and which is readily available at the Bureau of Labor Statistics.
The website gives you monthly data going back to 1948.
Available data are monthly, but I have only used the year-end numbers in the above graph.
The profile of employment across the United States was no doubt greatly different in 1948 than at present. In 1948 it was a lot less common to see both parents fully employed than was the case in 2000. Assuming that all families have at least one parent working full time, the peak value of about 67% in the late 1990s would suggest that 34% of families had two full-time working parents. Part-time workers are not counted in this statistic, and no doubt they were many.
The increase in participation from the early 1960s to the late 1990s reflects a major change in the structure of employment over the past two generations, with more married women entering the workforce.
In the lagged phase space we see two areas of attraction--the first in the 59% participation, and the second more recently, around 67%. We also note that the system has recently moved out of the higher area of attraction. The relative number of fully employed people is falling. How low can it go?
Below we see a topologically equivalent phase space constructed using the time-derivative method--plotting the participation rate against its rate of change, averaged over four years.
The same two areas of attraction appear, with what might be a third at around 64%. The current trajectory looks like it is heading there--so it doesn't appear that labor force participation will shrink much going forward.
Labor force participation increased during the period of falling interest rates after Volcker raised them so high in 1980. Their subsequent drop ignited a stock market bubble, which finally popped in the year 2000. The participation rate appeared to be heading lower--but intervention in interest rates inflated a bubble in real estate prices, "creating" lots of jobs in sales and construction--until it, too popped.
Based on the above graph, it looks as though most of the damage to employment of the bubble popping is over . . . unless . . .
Unless there is another bubble which we haven't considered. One that has been inflating since the early 1960s. One built on expansion of credit which has funded wars and expanded the economy to accommodate influxes of workers from family farms and young women wanting to take control of their own affairs.
Is this really logical? Did millions of jobs suddenly appear because there was this generation of young women who wanted to work?
One of the key features of a bubble is a story--a story that justifies the expansion, and which tells you that this time the expansion is not the inflation of a bubble. It looks to me that the happy story of women entering the workforce may be such a story--and there are still many millions more to leave the full-time labor force.
The website gives you monthly data going back to 1948.
Available data are monthly, but I have only used the year-end numbers in the above graph.
The profile of employment across the United States was no doubt greatly different in 1948 than at present. In 1948 it was a lot less common to see both parents fully employed than was the case in 2000. Assuming that all families have at least one parent working full time, the peak value of about 67% in the late 1990s would suggest that 34% of families had two full-time working parents. Part-time workers are not counted in this statistic, and no doubt they were many.
The increase in participation from the early 1960s to the late 1990s reflects a major change in the structure of employment over the past two generations, with more married women entering the workforce.
In the lagged phase space we see two areas of attraction--the first in the 59% participation, and the second more recently, around 67%. We also note that the system has recently moved out of the higher area of attraction. The relative number of fully employed people is falling. How low can it go?
Below we see a topologically equivalent phase space constructed using the time-derivative method--plotting the participation rate against its rate of change, averaged over four years.
The same two areas of attraction appear, with what might be a third at around 64%. The current trajectory looks like it is heading there--so it doesn't appear that labor force participation will shrink much going forward.
Labor force participation increased during the period of falling interest rates after Volcker raised them so high in 1980. Their subsequent drop ignited a stock market bubble, which finally popped in the year 2000. The participation rate appeared to be heading lower--but intervention in interest rates inflated a bubble in real estate prices, "creating" lots of jobs in sales and construction--until it, too popped.
Based on the above graph, it looks as though most of the damage to employment of the bubble popping is over . . . unless . . .
Unless there is another bubble which we haven't considered. One that has been inflating since the early 1960s. One built on expansion of credit which has funded wars and expanded the economy to accommodate influxes of workers from family farms and young women wanting to take control of their own affairs.
Is this really logical? Did millions of jobs suddenly appear because there was this generation of young women who wanted to work?
One of the key features of a bubble is a story--a story that justifies the expansion, and which tells you that this time the expansion is not the inflation of a bubble. It looks to me that the happy story of women entering the workforce may be such a story--and there are still many millions more to leave the full-time labor force.
You have to account for demographics too. Retired people have a much lower participation rate, so as the baby boomers age, the expected participation rate will drop even if each age group is participating at the same rate. I believe the data is available by sex and age group, which allows for much more detailed analysis of what is going on.
ReplyDeleteYou have something there that could be drilled down into. But there is data that suggests most of the job losses are for workers under 55, and that older workers are working longer (to support their kids who are having trouble getting started).
ReplyDelete