Arkady
President
Donald Trump, in his victory speech, said that unemployment may be over 40%. That is, of course, absurd. It's basically just redefining the term "unemployed" so that everyone 16 or older who doesn't have a job is unemployed -- including full-time students, retirees, the severely disabled, people here on visas that don't even let them work, full-time homemakers, and so on.
Specifically, the household survey says that 59.6% of the civilian non-institutionalized population 16 and older have jobs. That leaves 40.4% of them not working -- the mythical "unemployment" number. If you want to redefine the term that way, I guess that's just semantics and it doesn't matter. But the rhetorical punch of the talking point relies on people making an apples-to-oranges comparison, between that high figure and official unemployment rates of earlier eras that are much lower. If you redefine the term and do an apples-to-apples comparison, the view is very different. For example, although we now have 40.4% "unemployment," at the very height of our 1960s prosperity, in 1968, it was 42.2%. And back in the mid-1950s, "unemployment" was as high as 44.8%. Suddenly 40.4% sounds pretty damned good.
Trump may be a particularly shameless practitioner of this technique, but it's actually a widespread thing on the right. For example, other will refer to "real unemployment" as a stat that includes marginally attached workers (people who have given up looking), and those working part time for economic reasons (notwithstanding the silliness of a "real unemployment" stat that, by definition, includes employed people). When they do this, they likewise rely on people implicitly making an apples-to-oranges comparison between "real unemployment" today and standard unemployment in prior eras -- never addressing the fact that the decline in unemployment on Obama's watch is actually even greater if you do an apples-to-apples comparison using their alternative stat.
There's another game that right-wingers like to play, which is to select previously-obscure stats to harp on when the usual stats aren't telling them what they want to believe. For example, if unemployment is plummeting, focus on labor participation rates. If inflation is low, focus on some sub-set of consumer prices that's moving the other way (gas prices, if they're up, or the price of beef, or whatever you need to reach for). If every global survey of temperature shows an increase, focus on just satellite readings of a particular country. One of our posters here is particularly keen on this technique, though sometimes, when all the major indicators look good, it forces him to reach for arcana like the "Markit Flash US Composite PMI Output Index," to try to argue that everything's coming apart at the seams.
If you're interested in really understanding the world, rather than poring over data to try to find things that support your prejudices, it's a good idea to monitor your own use of stats. If you find that mostly you reference widely-referenced stats from mainstream sources (GDP, poverty, unemployment, median household incomes, the consumer price index, etc.), then it's a good sign. Even if you go for obscure or unusual stats, but you do so consistently, regardless of which way they're facing, and you have an argument for why they're more important than traditional stats, fine. But if you find yourself constantly having to reach for either invented statistics ("real unemployment," "misery index," Markit Flash Composites, etc.), and you pick them up and abandon them from month to month depending on what the latest numbers are saying, and you encourage apples-to-oranges comparisons, then stop and consider whether maybe this is a sign that you've gone hunting for stuff to bolster your prejudices, rather than material to inform your opinions.
Specifically, the household survey says that 59.6% of the civilian non-institutionalized population 16 and older have jobs. That leaves 40.4% of them not working -- the mythical "unemployment" number. If you want to redefine the term that way, I guess that's just semantics and it doesn't matter. But the rhetorical punch of the talking point relies on people making an apples-to-oranges comparison, between that high figure and official unemployment rates of earlier eras that are much lower. If you redefine the term and do an apples-to-apples comparison, the view is very different. For example, although we now have 40.4% "unemployment," at the very height of our 1960s prosperity, in 1968, it was 42.2%. And back in the mid-1950s, "unemployment" was as high as 44.8%. Suddenly 40.4% sounds pretty damned good.
Trump may be a particularly shameless practitioner of this technique, but it's actually a widespread thing on the right. For example, other will refer to "real unemployment" as a stat that includes marginally attached workers (people who have given up looking), and those working part time for economic reasons (notwithstanding the silliness of a "real unemployment" stat that, by definition, includes employed people). When they do this, they likewise rely on people implicitly making an apples-to-oranges comparison between "real unemployment" today and standard unemployment in prior eras -- never addressing the fact that the decline in unemployment on Obama's watch is actually even greater if you do an apples-to-apples comparison using their alternative stat.
There's another game that right-wingers like to play, which is to select previously-obscure stats to harp on when the usual stats aren't telling them what they want to believe. For example, if unemployment is plummeting, focus on labor participation rates. If inflation is low, focus on some sub-set of consumer prices that's moving the other way (gas prices, if they're up, or the price of beef, or whatever you need to reach for). If every global survey of temperature shows an increase, focus on just satellite readings of a particular country. One of our posters here is particularly keen on this technique, though sometimes, when all the major indicators look good, it forces him to reach for arcana like the "Markit Flash US Composite PMI Output Index," to try to argue that everything's coming apart at the seams.
If you're interested in really understanding the world, rather than poring over data to try to find things that support your prejudices, it's a good idea to monitor your own use of stats. If you find that mostly you reference widely-referenced stats from mainstream sources (GDP, poverty, unemployment, median household incomes, the consumer price index, etc.), then it's a good sign. Even if you go for obscure or unusual stats, but you do so consistently, regardless of which way they're facing, and you have an argument for why they're more important than traditional stats, fine. But if you find yourself constantly having to reach for either invented statistics ("real unemployment," "misery index," Markit Flash Composites, etc.), and you pick them up and abandon them from month to month depending on what the latest numbers are saying, and you encourage apples-to-oranges comparisons, then stop and consider whether maybe this is a sign that you've gone hunting for stuff to bolster your prejudices, rather than material to inform your opinions.
