Saving GM: the fallacy of composition of jobs

Topics: Regulation
29 Nov 2008

From: Ervan Darnell





Two of the obvious problems with the proposed GM bailout are that it
punishes productive companies in order to reward unproductive ones and it
encourages rent seeking instead of productive investment[1]. And,
it won't work in the long run anyway[2]. Others have covered these
problems well ([1] and [2] are worth reading, not just reference
footnotes for my argument).

Even if Krugman is right[2] (he begrudingly favors a bailout, but only
under the current circumstances, and not for long), when conditions
change and it no longer makes sense, you can bet Democrats will stop
listening. Perhaps this is why I have the temerity to disagree with
professional economists. While I won't dispute their technical
analysis as professionals, I think they undervalue the political risk
(where they are not professionals) of their proposals.

To try to add something to the discussion, I want to address two
fallacies I often hear:


1) Auto is 10% of manufacturing

So what? Banking is 100% of banking and GM is some tiny fraction of
the total economy. It's a gerrymandered argument. If I can
draw the line where I want, I can find a category that is 100% affected
or 0% affected. This argument thinks it's worse to lose
10,000 jobs in one place than to lose 10 jobs at 1,000 different
companies. Why? It's not. It just easier to see, that's
all. To first approximation, injecting $25G into GM must come from
somebody else doing something productive in order to pay taxes (in the
long run). That's $25G of induced unemployment elsewhere, just
scattered around, to save $25G worth of employment at GM.

But what about the knock-on costs of all that unemployment to parts
suppliers, etc.? It's the same argument. To first
approximation, the knock-on costs exists in the dispersed case too.
They are just more subtle, a restaurant worker here or a convenience
store worker there, loses their job because one less person in the
community shopped there because they lost their job at the insurance
company because of higher taxes to pay for the GM bailout, etc.

Yes, maybe there is an argument that concentrated losses cause
precipitous losses of certain individuals instead of spreading the pain
where it doesn't quite cause catastrophic losses (a marginal disutlity
argument or a tighter feedback loop). I'm not seeing that case
being made. Regardless, the simplistic argument about market
segments is too simplistic.


2) X million jobs at GM will be lost

Not really. Let's say Americans are going to buy 10M vehicles next
year. If they cannot buy those vehicles from GM, because it's
brankrupt, they'll buy them from somebody else.

Indeed, GM failing to get a bailout doesn't mean they are out of
business, it just means they go into chapter 11 bankruptcy. That's
a good thing because it lets them break their union contracts (which are
part of the problem, as GM has promised benefits to retired workers that
it no longer has the sales base to support, and the union pay scale
leaves them permanently uncompetitive with other manufacturers).
Maybe chapter 11 shouldn't exist; but for now it is the existing legal
alternative (indeed, I suspect that paying off the unions for their
political support is why Democrats are in favor a bailout that doesn't
make economic sense, instead of chapter 11).

But wouldn't consumers refuse to buy cars from a bankrupt company since
they don't get a warranty? I don't know what the ultimate
pscyhology would be, but it seems to me this doesn't make much
difference. If consumers think a chapter 11 GM won't honor its
warranties, then there is no reason to think a GM on government life
support will either. Besides, a warranty has some finite
value. A third party could offer an independent warranty that
promises coverage for anything on the GM base warranty were GM to fail to
honor its warranty. GM could then lower its manufacturer's warranty
period, lower the car price accordingly, and bundle the 3rd party
coverage, leaving the final car price and coverage the same as if they
weren't in trouble. If the government really must bailout GM, this
would be the way to do it (promise to fulfill the GM warranty so
consumers aren't artifically scared away).

Returning to the question of displacement, the "lost" jobs, are
not really lost. They will go to Ford or Nissan-USA(which hires
Americans to build cars in America) or some other car company.
Indeed, the best bailout plan there is for Ford is to let GM go
bankrupt. Yes, with the recession, fewer cars will be bought and
auto manufacturing jobs will be lost. But, those are going to be
lost regardless. Or, conversely, even if Congress thinks it has to
induce consumers to buy cars they cannot afford, there is no reason that
those cars should be purchased from the least efficient of the car
companies, GM. No matter which way you slice it, the lost GM jobs
are not lost, merely displaced.

Okay, not quite. Given fewer competitors, prices will rise a bit
(and need to!), and given some real preference by some consumers for GM
products over others, there will be less net auto buying utility.
But these differences are not 100% of GM's jobs, they are only a few
percent.

Or, you might argue that the jobs are sent overseas. But that's
just a more complex displacement argument, and not worth getting into
again here.

---------------------------------
[1]

http://www.cato.org/pub_display.php?pub_id=9787
[2]

http://krugman.blogs.nytimes.com/2008/11/16/cars/
[3]

http://www.washingtonpost.com/wp-dyn/content/article/2008/11/17/AR2008111703101.html



====================================================
Ervan Darnell,

http://www.kelvinist.com





_______________________________________________
Ragnar mailing list
Ragnar@ragnar.kelvinist.com
http://ragnar.kelvinist.com/mailman/listinfo/ragnar

Home