Off the Charts
I first became interested in how governments measure unemployment back in the early 90’s recession. I was unemployed and I couldn’t find a job and the same was true of many of my friends. Those who were able to find jobs were generally working farm below the level they had been in the past. And I’d read the unemployment rate numbers and it seemed to me they didn’t really reflect what I was seeing around me. So I decided to investigate. What I found out was that the unemployment rate didn’t measure exactly what I had thought it did.
Ian is a contributing editor at The Agonist..
The unemployed includes those people who don’t have a job, have looked for a job in the last four weeks, and who would take a job if it were offered. The unemployment rate is determined by dividing that number by the number of people in the labor force – which is equal to all the employed people + all the unemployed people.
It’s important to note that the labor force doesn’t include a number of groups. People serving in the military are not counted and neither is anyone who is institutionalized (which mostly means people in prison) isn’t either. Given the size of the US prison population (the largest per capita in the world) that’s not insignificant. Especially when you consider that blacks make up about 44% percent of the prison population (compared to 12% of the general population.) and blacks are much more likely to be unemployed than whites. In 2005 blacks had a 10.5% unemployment rate and whites a 4.4% rate. Consider that those who are incarcerated in general are more likely to have been unemployed (and more likely to be unemployed in the future due to the prejudice against ex-cons in hiring decisions) and you can see that keeping them locked up is not insignificant. (And this is one reason among many why you should be suspicious when people try to compare unemployment rates between countries.)
There are other caveats that tend to reduce measured unemployment. If, for example, you work for your family but aren’t paid, if your hours are over 15 you are considered employed, under 15 you aren’t considered part of the work force. So in one case – you decrease the unemployment rate, but in the other you don’t count as unemployed.
What all of this means is that the employment situation can be going to hell in a handbasket and the unemployment rate may not be going up very much. It may even go down.
So what I prefer to look at is not the unemployment rate – it is either the labor force participation rate which measures how many people are working, or still looking for work, and haven’t given up OR the employment to population ratio, which is simply how many people are employed as a proportion of the population. Because the working age population goes up every month, if the economy were to produce zero jobs, the job situation has actually gotten worse. Roughly speaking, to just stay even, the economy has to produce 150,000 jobs a month. Go take another look at the unemployment chart at the top, then compare to the labor force participation chart:
And compare to the employment ratio:
Note that the unemployment rate chart shows a significant improvement starting around July 2003 but the labor force and the participation rate charts show only a nominal improvement. People aren’t reentering the work force in large numbers – they’re staying out.
Another thing you’ll read a lot about is the "household survey" versus the "establishment survey."
The household survey is where the Bureau of Labor Statistics calls up households and asks them questions to determine if they are employed. The Establishment survey (also known as the payroll survey) is where the BLS calls up businesses.
There are two rules of thumbs about these surveys:
First – the establishment survey is generally considered more accurate. Apologists for the government in power will try and use Household numbers as a rule, because they tend to be better.
Second – there is a gap between the two surveys, with the household survey showing more people employed than the establishment survey. That’s because the household survey shows self employment, and the establishment survey doesn’t. In general if the gap is closing that’s a good thing – it means people are moving into payroll positions, and payroll positions on average, pay better than self-employment. When times are bad you’ll see the gap increasing as people do odd jobs and under the table work to keep themselves afloat. If the gap decreases while the labor force participation rate decreases at the same time (which happened recently) that’s a really bad sign – because it usually means you’re in a recovery, and it isn’t producing the sort of job growth that drives up wages and brings people into the job market.
(Note the distinctive heartbeat pattern. All the other charts used in this essay used seasonally adjusted figures – this one uses unadjusted figures. The heartbeat pattern is almost always there in unadjusted figures and shows graphically how seasonal our economy still is.)
So let’s present some rules of thumb to help you gauge how things are going:
- Forget the unemployment rate – look at the labor force participation rate or the employment/population ratio.
- Because the working age population of the US is going up every month, the job market must create jobs just to stay even. Roughly 150,000 jobs a month. If it creates less than that, consider things to have gotten worse.
- The Establishment survey is more accurate than the household survey. If the gap between the two surveys is decreasing that’s usually good. If it’s increasing, that’s usually bad.
I’ll deal with underemployment (ie. getting a worse job than your last one or than the one you’re qualified for) and with wages in a future post but for now, the bottom line about the Bush employment situation is that it’s worse than when he took office – as a percentage of the population less people are employed.