20 March 2012

Worse than the Great Depression?!

The Information Technology and Information Foundation, a DC-based think tank which routinely punches above its weight in policy debates over innovation, has a new report out today titled "Worse Than the Great Depression: What the Experts Are Missing About American Manufacturing Decline." The report has a lot of interesting data and arguments, which are worth unpacking in some detail.

However, in this post I simply want to address the overarching argument of the report that the experience of the past decade in the manufacturing sector has been -- as the title says -- worse than the Great Depression. This statement of the magnitude of the problem is thus the basis for arguing that policy actions are needed to protect manufacturing as a special sector in the economy. However, this overarching argument is hyperbole. It underscores the fact that ITIF has come to find itself on the wrong side of the innovation and manufacturing debate.

The figure at the top of this post helps to explain why the past decade is not particularly comparable to the Great Depression. The graph shows that while losses in manufacturing jobs were quite similar from 1929-1933 and 2000-2010, during the Depression the economy (inflation adjusted) contracted by 27% but during the latter period expanded by 18%!

One reason for this big difference is the much smaller role than manufacturing plays in today's modern economy. Almost by definition, manufacturing matters less to GDP and jobs than it once did. At the onset of the Great Depression, manufacturing was responsible for about 1 in every 3 non-farm jobs. Today, that number is more like 1 in 13.

The loss of manufacturing jobs is not unique to the US, as described by Mark Perry of the University of Michigan: "Australia's manufacturing/GDP ratio went from 22% in 1970 to 9.3% in 2010, Brazil's ratio went from 24.5% to 13.5%, Canada's from 19% to 10.5%, Germany's from 31.5% to 18.7%, and Japan's from 35% to 20%." These trends, I have and will argue, are the result of innovations that allow nations to get more output for less input.

A 2009 Congressional Research Service report (here in PDF and cited by ITIF) makes the case in more detail that the past decade is not comparable to the Great Depression:
There are substantial differences in the extent of unemployment during the Great Depression and the current recession. The unemployment rate rose almost eight-fold between 1929 (3.2%) and 1933 (24.9%). In contrast, it almost doubled between December 2007 (4.9%) and May 2009 (9.4%). At the peak of unemployment during the Great Depression (1933), one in four workers was unemployed, in contrast with fewer than one in eleven today. To approximate the pervasiveness of unemployment at the depth of the Depression, the number of workers without jobs would have to have totaled 38.6 million in May 2009, which is 24 million more workers than were unemployed this May (14.5 million).

Employers cut the total number of jobs on their payrolls much more deeply during the Great Depression than they have thus far in the latest recession. Between 1929 and 1933, employment on nonfarm payrolls fell by 24.3%, compared to 4.3% thus far in the recession. To approximate the relative extent of cutbacks that took place over the four-year period between 1929 and 1933, employers would have had to have shed 27.6 million more workers than they did between December 2007 and May 2009. In the goods-producing sector, 7.2 million rather than 3.0 million workers would have to have been laid off since the recession began to equal the relative impact of the four-year (1929-1933) decline. Within the goods-producing sector, construction companies would have had to have pared payrolls by 2.2 million more jobs than the 1.2 million positions cut through May 2009. Manufacturers would have had to have let go 2.5 million workers beyond the 1.8 million they displaced since December 2007 if the industry was in as relatively bad shape as it was in 1933.
The loss of manufacturing jobs over the past decade is not worse than the Great Depression, and that is a good thing, for the economy and for the future of American workers.


  1. It seems strange that you are comparing the worst four years of decline in the Great Depression to ten years of the recent recession, 7 of which are marked by growth. What would the graph look like if you changed the second period to 2007-most recent data?

  2. Roger, as the lead author of the ITIF report, let me make 2 comments. The report never says that the economy of today is worse than the Great Depression. It just says that the rate of job loss in manufacturing is worse. not that is central to my response, but your chart uses nominal GDP dollars, not inflation adjusted.

    Finally, your discussion of other nations manufacturing share of GDP is not really meaningful. First, it looks like Perry is using nominal dollars, which is misleading. If manufacturing productivity is higher than non-manuf productivity, then nominal share of GDP is more likely to fall, even though in real, inflation adjusted terms it does not. our report explains this. Second the real issue is whether its declined more in some countries than others because of loss of competitive position. And this is clearly the case in the US (and Candada) again as the report shows. The problem with this debate is that too many people have become attached to two stories 1) its all productivity that leads to job loss; and 2) manuf is declining everywhere so our decline is normal. If productivity is partly the reason it doesn't mean its the full reason. Likewise, just because manuf is declining in some nations (and growing in others0 it doesn't mean its all due to declining manuf demand. In fact, in the US none of it is due to demand issues.

  3. -2-Rob Atkinson

    Thanks for the comments! A few replies on the data first:

    1. Yes, there was a spreadsheet snafu in the recent GDP, now fixed (Thanks!). It does not alter the argument, however.

    2. I do not understand your inflation/non-inflation comment wrt Perry's numbers. By definition, an inflation adjustment alone will not make a difference on calculating the share of the economy. Your point must be about something other than inflation.

    3. On the "real issue" -- your use the the terms competitiveness and productivity fails to recognize that these concepts are intimately inter-related. If Foxxconn in China can assemble Ipads at lower overall costs than (say) Acme in Illinois, then this competitive advantage is at the same time an improvement in productivity.

    Bottom line on the point of this post, to claim that the recent loss of manufacturing jobs is worse than the Great Depression is a bit like saying that 2012 NY Giants had a worse Super Bowl than the 2011 Pittsburgh Steelers because the Steelers scored 25 points and the Giants a mere 21. Of course, the Giants won their game and the Steelers did not.

    Data only matter in a context.


  4. Roger to respond to your response to mine: the real issue is that the only valid measure of manufacturing share of an economy over time is to use "real value added as share of gdp" in other words, how many "Cars, couches, and lightbulbs were produced controlling for quality. Perry doesnt do that.

    the foxcconn example has nothing to do with productivity. If they can assemble at a lower cost through lower cost this is not higher productivity. but the point, that almost everyone who rejects our argument seems to make is that by defending manufacturing, we are defending it all equally. of course not. with global division of labor some things will move to other nations. but as Germany shows, restructuring does not have to mean decline. In the US it has meant decline.

    finally, to claim that the loss of 1/3 of US manuf jobs and the fact that we created no net jobs in the 2000s are unrelated is to missing missing the elephant in the room.

  5. -1-Chris Chambers

    Those years were the ones used in the ITIF report, so I also used them here. Thanks!

  6. -4-Rob Atkinson

    I am baffled about your claims regarding Germany. Here is manufacturing employment in Germany:


    Sure looks like decline to me. What am I missing? Thanks!

  7. ". If they can assemble at a lower cost through lower cost this is not higher productivity. "

    On the contrary it is exactly higher productivity.

  8. The Great Depression didn't happen overnight, but it was a much more sudden collapse than the recent loss of manufacturing jobs. By the year 2000, many manufacturing plants that had been open for decades had already closed. In the last ten years, there have been far fewer jobs to lose than there were 30 years ago, so of course the numbers can't compare to the Great Depression.

  9. what happened to China's manufacturing/GDP ratio? it has stayed flat or also declined?

  10. "with global division of labor some things will move to other nations. but as Germany shows, restructuring does not have to mean decline. In the US it has meant decline."

    I think you need to read your report more carefully. ;-)

    If one goes to Figure 5 of your report, one sees that the decline in U.S. manufacturing jobs was heaviest in:

    1) Textile mills (68 percent)
    2) Apparel (67 percent)
    3) Leather and allied products (59 percent)
    4) Textile product mills (48 percent)
    5) Furniture and related products (47 percent)
    6) Wood products (45 percent)

    For those top six percentage-losers in U.S. manufacturing, Germany was not a signficant producer in *any* of them.

    So of course the U.S. lost jobs in those sectors, but Germany didn't. That's simply because Germany had no jobs in those sectors in the first place.

    P.S. I can't get off this discussion without recalling some of the best lyrics from The Boss (and that's really saying something, since he's one of the great lyricists of all time):

    "Now Main Street's whitewashed windows and vacant stores.
    Seems like there ain't nobody wants to come down here no more.
    They're closing down the textile mill across the railroad tracks.
    Foreman says these jobs are going boys and they ain't coming back to
    Your hometown, your hometown, your hometown, your hometown."

  11. #9 John

    According to The Economist Pocket World in Figures 2012 ed. China's GDP was 34% manufacturing (part of industry46%), Agriculture 10% and services 43%

    Employment industry 25%, agriculture 41% and services 34%

  12. #9 John

    I found the 2003 ed.
    Agriculture 16.3
    industry 51.3
    services 32.1

    agriculture 65
    industry 22
    services 13