Make-work

What does “emergency” mean here? Does this mean employees of FEMA? Or firefighters and ambulance drivers? Or something else? I would also assume that most of these jobs of either type would either count as “government” or “civilian”.

Kenny – it’s people on make-work schemes.

“will be net out” -> “will net out”

Moreover, from a theoretical perspective, “emergency employment” serves two economic functions: first, it’s a very effective Keynesian stimulus, because it pulls people off the unemployment lines and gets them spending; second, it increases overall economic growth directly, because it taps into the unused labor power of unemployed workers – which is usually worth more than the wages concerned.

(er, increases in the blue line should reduce the green line. but you knew what I meant.)

I have one methodological criticism: the use of “Millions Employed” fails to compensate for population growth. (Or shrinkage, if the Depression was bad enough to kill millions of people.) Does the chart look noticeably different with a y-axis of “Number employed as % of U.S. population”? (Or “of U.S. potential workforce”, if there’s a sufficiently uncontroversial definition and numbers available.)

I would expect not, if the population is fairly stable or growing in a smoothish way, but it’s always nice to have the correct denominator when you’re looking at changes in ratios.

Ah, then I guess population change wasn’t a major factor.

I forgot to mention this before, but – it’s interesting that this is *yet another* measure of the economy showing that it got steadily worse from 1929-1932, *just happened* to turn around right as FDR took office and implemented his completely useless and/or counterproductive programs, then improved year by year until 1937, reversed in 1938, and then went back to growing again.

It’s almost as though something with an actual causal effect on the economy happened right around 1932-1933, but all the serious right-wingers tell me FDR had nothing to do with it, so I guess the timing is just a big coincidence.

On a more serious note, from someone who is neither an economist nor a historian of the Depression – so what did happen in 1938, anyway? Policy change? (Is it just my imagination, or does the “Emergency” category of employment shrink slightly in 1937? It’s hard to tell when it’s piled on top of other variables…) Natural disaster? The business cycle? Something else?

The point of the graph is [to] show that even if you accept Tabarrok et al.’s position that WPA-type stuff is not “real” work, the claim [the that New Deal didn’t improve employment] is still false.

Lemuel, I’ve been following this whole debate pretty closely (go over to the comments at Edge of the American West, for all of the entries on this issue). The problem is, the graph itself doesn’t prove that the New Deal improved employment. Without a proven mechanism, pointing to the graph is just argument from correlation, it’s post hoc post propter hoc.

To a hardcore New Deal Denialist, the graph would only “show” that the New Deal didn’t do enough economic damage to preclude some “natural” (laissez faire) rebound in employment. Then they move the argument on to (or back to) the old assertion that the New Deal prolonged the Great Depression. Next thing you know, lunch is over. No mind has been changed.

And that’s how it goes with these folks. To a certain mentality, all free market failings not attributable to evil government intervention can be traced back to a loss of faith in the Magic of Markets. (Except for market failures like bubbles, which are held to have some actual benefits, and anyway should only be left to themselves to cure. Which they will. Cuz that’s how the Magic of Markets works, dummy! By magic!) In this view, government action is evil wherever it enables or encourages any such loss of faith.

In fact, I used to partake of this mentality (long sordid story, won’t get into it here.) The funny thing is, I can’t even tell you why I did. It sort of reminds me of some observations due originally to Piaget — in certain reasoning/perception tests, children at certain ages consistently get it wrong, then at a later age are able to get it right, while finding hilarious the suggestion that they ever saw it any other way. The notion that the Paradox of Thrift, combined with deflationary expectations in the wake of bubble overinvestment, could result in indefinite stagnation, requiring an external demand shock for correction, seems pretty obvious to me now.

Could it be that the possibility is only obvious to people who are smarter than the average bear? My intellectual vanity whispers yes, but I don’t think that’s it. With a well-designed classroom treatment (including economic simulation games), couldn’t the whole chain of reasoning and how it works in real life be put within the grasp of junior high school students? Maybe only reason it isn’t standardly taught is ideological: that it’s “controversial”? Y’know, that it’s just a theory? Like Darwinism?

You’re correct that it is correlation, but it’s very strong correlation, especially the way it tracks the spending cut to spending boost flip.

I’m hip, I’m hip. (I make the same point, in fact, in my comments on this graph over at Edge of the American West.) All I’m saying is: you’d still get arguments from True Laissez Faire Believers, so it’s better to watch how you say things. Don’t give them easy openings.

For example, Kieran might have meant his “Make-work” blog heading facetiously. If so, however, without scare quotes, it looks like at least one commenter here took it at face value. When you say “shows” (interpretable as “proves”) it’s also going to get you tripped up with the New Deal Denialists. Whatever their limits in reasoning, they are still possessed of certain (myopic) logic skills. Skills they apply selectively, of course: recall that an awful lot of the ruckus was over whether it was fair to say someone is lying when they assert that unemployment “remained” at around 20% in 1937/28. Some say it’s still true (therefore honest), in some sense, provided you’re not so frickin’ nitpicky about what you mean by “remained”. Call me a nitpicker: if you say something “remained”, you’re saying it didn’t move elsewhere after somebody did something or after something happened. Whatever: the Faithful will give each other a pass on when one of them uses this “remained”, but not be so charitable about their opponents’ “shows”. And if you call it “make-work”, they’ll look up the word just to be sure, then accuse you of frankly and impenitently admitting to being in favor of “wasteful government spending.”

Good luck on your counterfactual modeling effort, by the way, I’m all in favor this kind of thing. Of course, if you get further confirmation, you’ll probably only get a “models are bunk” response from the kinds of people who knee-jerkedly blame the current financial crisis (and/or “global warming hysteria”) on Those True Modeling & Simulation Believers (even though, in the case of modeling mortage backed securities, much of the more speculative trading was actually in defiance of models that suggested traders should be more conservative.) Over the next few years, maybe we’ll get more confirmation and better calibration on modeling the Great Depression. It might even convince a few more people . . . .

Ah, good thought, Attewell. I suggested the very same approach, and even saw it implemented, before they gave me the boot from my role as High Commissioner of the NDCTRC (scroll/search for it in comments). You might try a few modeling runs under their key assumptions, just to balance the study with scenarios more charitable to the Denialists.

Chris 11.20.08 at 1:00 am

Ah, then I guess population change wasn’t a major factor.

“I forgot to mention this before, but – it’s interesting that this is yet another measure of the economy showing that it got steadily worse from 1929-1932, just happened to turn around right as FDR took office and implemented his completely useless and/or counterproductive programs, then improved year by year until 1937, reversed in 1938, and then went back to growing again.”

There’s an engineering term used at Ford, ‘ABA experiment’, meaning start in condition A (Hooverism), switch to B (New Deal) and then switch back (1937-8), to show that you can induce and then remove a condition. This is far better data that most of what economists analyze, so it says something about any economist who refuses to accept it.

Michael Turner:

“Lemuel, I’ve been following this whole debate pretty closely (go over to the comments at Edge of the American West, for all of the entries on this issue). The problem is, the graph itself doesn’t prove that the New Deal improved employment. ”

Actuall, it does, from the viewpoint of most social science. It is true that this wasn’t done by the Paratime University Department Social Experimentation, with a timeline split induced to run parallel New Deal and No New Deal alternate universities.

However, in fields (like the frakkin’ social sciences) where time series data are the norm, it’s good proof.

“Without a proven mechanism, pointing to the graph is just argument from correlation, it’s post hoc post propter hoc.”

Two points – the mechanism (I’m not sure what you mean by ‘proven mechanism’) is given by Keynesian economic theory. Second, it;’s not just one change (implementation of the New Deal, with no late ’30’s partial pullback), it’s multiple changes (implement, partial pullback, restore). That is a very good proof.

the mechanism (I’m not sure what you mean by ‘proven mechanism’) is given by Keynesian economic theory

Yes, I know. Jeez. When I mention my pro-Keynesian comments at links, you might actually go read what’s there. Or maybe just read me more carefully here. I’m not arguing the Denialist case. I strenuously, repeatedly, argue against it. But weaknesses in arguments against it are still weaknesses in arguments against it, and I’ll point those out too.

Seemingly unnecessary caveat: when I say “proof” here, I mean it in the loose sense of “scientifically proven”, which should be understood as “all other plausible theories have been disproven or shown to be very implausible.”

I don’t agree with your assessment of standards of proof in the social sciences. As a blanket assessment, at least, it sounds unfair. But I’ll leave it to the real social scientists at CT to correct your view, if necessary.

Chris:

No, it wouldn’t be hopelessly naive; I’d just say “temporarily suspended” rather than abandoned.

As to the larger question – I think the way to tell if it’s working is to set a clear set of macroeconomic guideposts that you’re trying to hit as a way to focus Keynesian economic planning on a set of key variables. Just thinking off the top of my head, but they might have been something like unemployment at 3% or below (Beveridge’s estimate of full employment), growth of 5% or better, inflation below 3%, or something along those lines.

The important thing is to ratchet down spending at a slower pace than the economy is improving – FDR’s problem wasn’t that there wasn’t economic recovery going on, but that it was too fragile to absorb a sudden 100% cutoff of stimulus. Had he decided to cut back at a slower rate, it might have worked. (Probably not, given that the unemployment rate was still high at 9%)

“Aside from the logical paradox you’ve stated . . . .”

What logical paradox? We’re talking science here, not mathematics. As in mathematics, you can prove a scientific theory wrong with just one counterexample, (theory: no mammal lays eggs; disproof: the platypus), but you can’t prove a scientific theory right in the logical sense that you can prove a mathematical theorem right.

Science is a process that deals only in falsifiable statements, and results only in an accumulation of falsifiable statements that have successfully resisted attempts to disprove them. Very often there’s a proposed mechanism, but not always. If you tell an astronomer, “the heliocentric theory only looks correct, because God (or the Flying Spaghetti Monster) arranges the universe to look that way to us,” an astronomer will shrug and say, “science can’t prove you wrong; on the other hand, nobody can. We’re scientists, we don’t play that game.”

Barry, earlier you appeared to be unaware that the Great Depression did not kill millions of Americans. (Note that WW II killed fewer than half a million Americans.) Are you sure you’ve done enough background reading to contribute productively and coherently to this discussion?

As in mathematics, you can prove a scientific theory wrong with just one counterexample, (theory: no mammal lays eggs; disproof: the platypus), but you can’t prove a scientific theory right in the logical sense that you can prove a mathematical theorem right.

Part of science is about what happens. For that part you have to actually look and see what happens. And part of science is about what you call things. Definitions. That part is somewhat arbitrary but not completely arbitrary, and it’s more like mathematics. If you say that livebearing is one of the things that mammals have to do, part of the definition, then you’ll say that platypuses are not mammals. You might wind up with a new name for it. But if you decide that it’s a mammal then you’ll accept the egg thing.

It’s a definitional choice, not a proof.

In math you start with choosing axioms and then you find out what results from those axioms. In biology you start with what happens and you keep looking for definitions that better classify it, and also you look for axioms that might fit.

Barry, earlier you appeared to be unaware that the Great Depression did not kill millions of Americans. (Note that WW II killed fewer than half a million Americans.)

The way I read what he was saying, he saw somebody else imagining how things might have gone with a lazy fare approach, and they assumed things would turn up peaches. He wanted to imagine the same thing with a different slant. He supposed that millions of people would die from the failed economy, and then the question would be how we should count that in terms of unemployment. People who die don’t go on unemployment rolls afterward, but if you’re comparing outcomes maybe they ought to. Like, an economy where unemployed people are rounded up and killed would officially have an unemployment rate close to zero, but you wouldn’t quote their unemployment statistic like it was a *good* thing…..

I think that the best summary of my position can be expressed (and I have before) concerning Michael Turner’s comment #15, above:

“To a hardcore New Deal Denialist, the graph would only “show” that the New Deal didn’t do enough economic damage to preclude some “natural” (laissez faire) rebound in employment. Then they move the argument on to (or back to) the old assertion that the New Deal prolonged the Great Depression. Next thing you know, lunch is over. No mind has been changed.

And that’s how it goes with these folks. To a certain mentality, all free market failings not attributable to evil government intervention can be traced back to a loss of faith in the Magic of Markets. (Except for market failures like bubbles, which are held to have some actual benefits, and anyway should only be left to themselves to cure. Which they will. Cuz that’s how the Magic of Markets works, dummy! By magic!) In this view, government action is evil wherever it enables or encourages any such loss of faith.”

This could (perhaps unfairly) summarized as: ‘No matter what proof, there will be a set of people Who Just Won’t Accept It.’. For example, in the 150 years since ‘The Origin of Species’, science has produced a vast range of evidence supporing evolution, some of which would have been considered to be sheer magic to the scientists of Darwin’s age. But creationism just keeps plugging along (albeit not in science, where evolution has been proven
to the satisfaction of 99.9% of people involved). The disbelievers merely make up reason after reason, frequently recycling older, discredited reasons, once they’ve cooled a bit in obscurity. More evidence doesn’t persuade, and even proving that the ‘experts’ in creationism have to be lying doesn’t persuade – and that’s a commonsense approach, building on the layman’s experience that the proven liar shouldn’t be trusted.

And in terms of economics itself, we’ve in the middle of watching a Lesser Depression (if we’re very, very lucky), which sure as h*ll wasn’t predicted by the majority of Econ Nobels; the bulk of their predictions are probably now laughable. But do you really expect to walk into Chicago in ten years, and find anything other than hardcore right-wing politics taught there?

At some point, you’ve just got to call a crank a crank, and a liar a liar.

Note: CT did something perverse to my fonts; I’ve tried to fix them, but it’s hard to fix invisible things.

It was confined to preview; it must be a preview bug. Of course, I had no idea that preview was lying to me.

“If a plausible theory has been shown to be very implausible, it’s no longer plausible. This wasn’t a serious dig, but a gentle nudge at a logical flaw (and I’m not claiming to be above such errors on my own).”

Yeah, but it’s obviously (I thought) implied in what I said that the theory you’re claiming is proven is not one of the very implausible ones (or one of the disproven ones). Obvious enough to me, anyway.

In any case, Barry, you still need to substantiate your claim that, in the standards of proof for the social sciences, mere correlation is enough. I’m not a sociologist, but I don’t see such looseness in the sociology I’ve looked at.

As for how I got the impression you thought the Great Depression might have killed millions, it was Chris’s “Ah, then I guess population change wasn’t a major factor”, which you quoted later in such a way that it looked to me like it was your response to Chris, not a quote from him. Sorry for the mistake, but you might consider cleaning up your quoting style if you don’t want to be misunderstood this way. Note how you quote me in #17, above: properly, the second paragraph you use should also lead off with a quote. Better yet: blockquote the whole thing, or put it in italics, the way others here do.

As long as I’m here:

And in terms of economics itself, we’ve in the middle of watching a Lesser Depression (if we’re very, very lucky), which sure as h*ll wasn’t predicted by the majority of Econ Nobels;

They weren’t predicting a Lesser Depression, and I don’t think any are saying that now. At worse, I’ve read that they expect a long, deep recession. Better policy responses, starting last summer, might have headed off even that prospect. Better policy responses even this summer might have made some difference. It’s very hard to predict a recession (Brad Delong maintains you can’t know for sure until you’re in one, which is hardly “predicting”, is it?), and it’s also hard to predict policy responses.

. . . the bulk of their predictions are probably now laughable.

Has it occurred to you to go look at the bulk of their predictions?

But do you really expect to walk into Chicago in ten years, and find anything other than hardcore right-wing politics taught there?

Austan Goolsbee, Obama’s main econ advisor, is on the faculty of U. of Chicago. But their econ dept could lose all its liberals, and would still not exemplify “hard-core right-wing politics”. In economics, that’s usually “end welfare”, “total laissez faire” and “go back to the gold standard.” Even Milton Friedman was in favor of a guaranteed national income, and was a monetarist, or “fiat money” heretic, to the gold bugs. I’m sure you can even find him guardedly endorsing certain interventions here and there.

You’re kind of overheating here, Barry. I repeat my suggestion: some background reading, first?

Me: “But do you really expect to walk into Chicago in ten years, and find anything other than hardcore right-wing politics taught there?”

MT: “Austan Goolsbee, Obama’s main econ advisor, is on the faculty of U. of Chicago. But their econ dept could lose all its liberals, and would still not exemplify “hard-core right-wing politics”. In economics, that’s usually “end welfare”, “total laissez faire” and “go back to the gold standard.” Even Milton Friedman was in favor of a guaranteed national income, and was a monetarist, or “fiat money” heretic, to the gold bugs. I’m sure you can even find him guardedly endorsing certain interventions here and there.”

No, ‘go back to the gold standard’ is far beyond right-wing economics, into people who use ‘fiat currency’ as a curse, rather than as a technical term. It’s like equating ‘hard core leftwing professors’ with actual unreconstructed Khmer Rouge supporters. The presence of Goolsbee impresses me no more than the presence of Norm Ornstein (sp?) at AEI; I’ve come to call it the ‘Sodom and Gomorrah principle’ (keep one honest man around, so God don’t smite ya).

In terms of ‘hard-core rightwing economics’, perhaps I should clarify – do you expect to go into Chicago in 5 or 10 years, and find econ professors there praising the regulatory regime which (hopefully) the Obama administration will implement? Or will they be saying that it’s unnecessary, and too restrictive, and not needed in ‘these modern times’? [amazing, isn’t it, how ‘modern times’ call for changed measures, from people who read Adam Smith?]