Technological Progress in the 1930s
Via Tyler Cowen, economic historian Alexander Field makes the case that the 1930s was the decade in which we saw the most technological progress:
Because of the Depression’s place in both the popular and academic imagination, and the repeated and justifiable emphasis on output that was not produced, income that was not earned, and expenditure that did not take place, it will seem startling to propose the following hypothesis: the years 1929–1941 were, in the aggregate, the most technologically progressive of any comparable period in U.S. economic history. The hypothesis entails two primary claims: that during this period businesses and government contractors implemented or adopted on a more widespread basis a wide range of new technologies and practices, resulting in the highest rate of measured peacetime peak-to-peak multifactor productivity growth in the century, and secondly, that the Depression years produced advances that replenished and expanded the larder of unexploited or only partially exploited techniques, thus providing the basis for much of the labor and multifactor productivity improvement of the 1950’s and 1960’s.
I think a phrase like “the most technologically progressive” period is hard to define properly. But there’s no disputing the fact that there was substantial technological innovation during the Depression (refrigerated trucking as we know it, to cite just one example, arose during this period) and a ton of productivity growth. You can tell about the productivity increased by the fact that GDP had fully recovered to its 1929 peak by 1936, was clearly higher in 1937, remained above ‘29 levels throughout the 1938 trough, and then was higher still in 1939 and 1940 even though the unemployment situation remained bad for much of this period, and absolutely terrible during the ‘37-’38 recession-within-a-depression:
I assume that part of the story here is that the Roosevelt administration implemented labor market policies that had the effect of pushing real wages up in what they thought was an anti-deflation measure, rather than letting them fall in a way that would have encouraged more employment. This should have given employers and workers incentives to try very hard to make labor-hours as productive as possible.
26 Responses to “Technological Progress in the 1930s”
Register
RSSLogin
Topic Cloud
ArchivesBlog RollAbout Matt YglesiasContact Me
Donate
March 12th, 2009 at 11:42 am
Where are you getting your views on unemployment during the depression ? In particular are you counting WPA workers ? The things they produced are counted in TNP but the old (no longer official) unemployment series counted them as unemployed.
If they are counted as employed the horrible unemployment during the depression within the depression (38) was 12.5%, which is horrible, but half the level of unemployment when Roosevelt was inaugurated.
http://tinyurl.com/48j7cj
You will be surprised by productivity growth if you count production as production and the efforts of the workers as unemployment won’t you ?
March 12th, 2009 at 11:43 am
March 12th, 2009 at 12:03 pm
You’re looking at the wrong graph. You should be looking at GDP/Person Employed (although it’s more complicated than even that). GDP will naturally grow as a result of population increases; I’d bet the US population was growing very rapidly during the 1930s as a result of large gains in sanitation improvements.
This is why economists use the phrase factor productivity. Output is usually given as (lambda)K * L = C, where K is capital, L is labor, and C is output, and (lambda) is factor productivity. Output can grow via increases in capital, labor, or via increases in factor productivity.
Showing a graph of GDP growth doesn’t show what the determinant of growth was – could likely have just been from an increase in labor as anything else. You need to run models to sort out exactly what the causal factor of increased output was.
This is, Matt, why Tyler Cowen gets paid the big bucks and you do not.
March 12th, 2009 at 12:16 pm
DTM, what was the increase in working-age population versus the increase in unemployment?
Also, one would have to know where the job losses are located. Were most of the lost jobs in low productivity areas like farming? And were gains in employment in relatively high skilled sectors? So that total employment goes down, no productivity growth occurs, but due to changes in industrial patterns, total value of output goes up?
It is hard to sort this out.
March 12th, 2009 at 12:19 pm
It’s working age population that you should care about. I don’t think there was a lot of child labor during the Great Depression. So the relevant demographics should start at about 1912 through 1922 (maybe a little later because I suppose people could have started working at 16). Also, increases in standards of medical care may have allowed a greater percentage of the workforce to remain in the workforce for longer.
**
Ultimately, I’m sure the paper is correct. But what I’m saying is that you have to take a lot more into account than Matt’s simple graph to get anywhere. Like, for example, a multi-factor analysis of productivity growth using advanced statistical software.
March 12th, 2009 at 12:20 pm
“This is why economists use the phrase factor productivity. Output is usually given as (lambda)K * L = C, where K is capital, L is labor, and C is output, and (lambda) is factor productivity. Output can grow via increases in capital, labor, or via increases in factor productivity.”
The equation is actually fairly obvious, doesn’t illuminate that much and may even obscure more than it enlightens. Matt’s admittedly literary bumblings in economic history may even be more worthwhile than many an economist’s over-confidence in models that turn out not to be very good ones.
March 12th, 2009 at 12:23 pm
burritoboy – Okay, that growth model is of course an oversimplification for the purposes of comments on Matt’s blog – there are tons of them out there, most of course being variants of the Cobb-Douglas function, but there are plenty of obscure ones.
My purpose was to point out that even in the case of the braindead simple growth function model, Matt’s graph falls apart at proving his point. Under more complicated models, it obviously has even less illustrative value.
March 12th, 2009 at 12:43 pm
Crap, lost my post anyway, I don’t understand why this meme persists:
I assume that part of the story here is that the Roosevelt administration implemented labor market policies that had the effect of pushing real wages up in what they thought was an anti-deflation measure, rather than letting them fall in a way that would have encouraged more employment.
Anyone responsible for hiring in a business(as I was) knows that the only variable driving employment is how well your sales are going. If I only need 20 employees, based on current conditions, I’m NOT going to hire new employees if the wage rate is cut by half, I’m just not. Furthermore if business conditions require me to CUT 10 employees, I’m not going to keep these employees if their wages are cut, no matter how low they go.
March 12th, 2009 at 12:54 pm
The first thing that comes to mind is when you cut employment, you (usually, assuming seniority rules and the like aren’t extremely pathological) cut the least productive workers first. Therefore productivity (output in widgets per worker, even controlling for increased capital per worker) goes up automatically.
March 12th, 2009 at 1:03 pm
wml – This is true, but economists account for that already. Labor in most growth regressions is usually considered as something more like (labor x skill), so you’re not just counting bodies. Therefore, firing lower skilled workers will decrease a small enough of L to balance with the decrease in C.
Thus, the (lambda) of factor productivity will actually represent increases in technologically-based productivity increases (of course, it’s more complicated than that, but from a 30,000 foot perspective it’s usually how it works out).
March 12th, 2009 at 1:07 pm
If I may illustrate the thesis with a popular bit of anecdotal evidence: just compare the 1929 Ford Model A to the 1941 Ford. You barely need to pop the hood to know that Ford made a huge leap forward in those 12 years.
March 12th, 2009 at 1:13 pm
Maybe the really interesting point here is that, even after huge amounts of unproductive outmoded investment were bled out of the system, and huge amounts of new improved productive capacity were added, it took the kick-start of WW II to bring the productive capacity on line.
When you look at figures for actual companies, you often find declines of 90% between 1929 and 1934. In agriculture, the 20s were disastrous as the substitution of the tractor for the horse freed millions of acres for market production, ruining the farmers. Whatever the technological advances may have been, there were simply no markets to support them.
I’ve long found this decade of the 30s to be of interest because it seems intuitively obvious that it’s a hinge-point of the transition from coal to oil and hydro-electric. Now we’re very close to another point at which we change our sources of energy.
Of course, older folks like myself might just figure, why bother? Frying the planet like an egg won’t happen while I’m alive, so why rock the boat?
Wouldn’t that be fun?
March 12th, 2009 at 1:21 pm
DTM – I guess it just gets my dander up because people make this mistake all the time.
I can’t tell you how many people I’ve seen suggesting some kind of productivity miracle in China over the last 30 years. Yes, there has been an increase in total factor productivity that cannot be sneezed at. But neither can increases in K and especially L.
Understanding that GDP growth is a multi-factor product of at least some combination of increases in labor (not just bodies, but also labor skill levels), total capital available, and productivity would go a long way toward understanding why it isn’t especially shocking that the US has had stronger GDP growth than Europe, yet they end up with higher GDP/capita growth (in PPP terms).
March 12th, 2009 at 1:39 pm
the US has had stronger GDP growth than Europe, yet they end up with higher GDP/capita growth (in PPP terms).
As reported by the OECD, growth in GDP/capita (in PPP terms) in the United States and the EU15 was virtually identical between 1970 and 2006. Over the more recent period from 2001 to 2006, growth has been higher in the U.S. than in the EU15.
March 12th, 2009 at 2:21 pm
Andrew at 6:
Re: where unemployment losses were, you can find them in the section of the 1940 census titled: “Comparative occupation statistics for the United States, 1870 to 1940, a comparison of the 1930 and the 1940 census occupation and industry classifications and statistics: a comparable series of occupations statistics, 1870 to 1930; and a socio-economic grouping of the labor force, 1910 to 1940..”
March 12th, 2009 at 2:25 pm
the 1930s were the last blowout of Norman Angell’s world, and it represented many of the last gains of the victorian age’s globalization. Cross-fertilization was at a peak in the sciences and engineering. In general, the revolutions that happened in the 1930s, to me, was probably a product of increased urbanization and communications.
The innovations held on and were fully exploited because of world war though. If *we* had a *real* war tomorrow, we’d start mining much of our substantial collections of knowledge and putting them into working ideas as well. In general, the 1930s were full of experiment-lab conflicts like the Spanish Civil War…
March 12th, 2009 at 4:27 pm
Having charts/grpahs that don’t pop to a legible size for all the words makes the FSM lose a noodle.
March 12th, 2009 at 8:13 pm
Why is this at all surprising, Matt?
Hell, we went from this to this in fifteen years.
March 13th, 2009 at 2:32 am
Between the Hoover Dam, the Tennesee Valley Authority and hundreds of smaller projects, government vastly increased the supply of electricity. So the figures don’t surprise me, but I do find it irritating how conservatives remain in denial that government did good things, and remain opposed to trying similar things today.
March 21st, 2009 at 9:37 pm
Buy Viagra…
http://url.edna.edu.au/4bbN Buy Viagra…
March 26th, 2009 at 1:12 pm
Buy Propecia Online…
http://www.folkd.com/user/buypropeciaonline Buy Propecia Online…
April 6th, 2009 at 1:50 am
hpwl nqcj huzc fszpn znrgex jtlodvr ziufvcnj