Monday, December 22, 2008

On the Division of Labor Numbers

According to the Washington Post (Hat Tip: Russ Roberts), Obama plans on spending $670 to $850 billion dollars to save or create 3 million jobs (in an economy that destroys and creates 30 million jobs annually). Let's ignore the absence of evidence on government spending to create the growth necessary for jobs, and instead just do the division:

$670 Billion/3 million = $223,000 per job
$850 Billion/3 million = $283,000 per job

So the cost per job will range somewhere in the neighborhood of $250K, and I think it is fair to assume that these jobs will likely pay less than that on average. According to the Census, in 2007 there were 75.3 million people in the U.S. between the ages of 18 and 64. Let's assume that the burden falls entirely on them, in which case the average person would be individually paying $8,900 to save these jobs.

Even if we actually believed their job saving claims, what about this proposal could possibly seem appealing? It feels right to leave with a quote from Bastiat:

"The State is the great fiction through which each of us tries to live at the expense of all of us."~French economist Frederic Bastiat


Suzie said...

Your article is questioning whether this $9,000 input from an individual is worth the output of a new job. What if some of these "new or saved" jobs are contracting public works projects such as bridge upgrades? These projects have been deferred onto future generations and with a need to boost an economy why not hire people to complete them now? Our $9,000 would be going to both a job and the job output that collectively we benefit from.

Justin M Ross said...

Suzie: Thanks for the question, it's a common argument that needs to be addressed...

The value of their production is already capitalized into the worker's income (i.e. worker's are paid their marginal revenue product). Therefore, it would be incorrect to count as social benefits both worker income from the employment and the value of output produced because it would be double counting.

Remember also that the $9,000 per person would not otherwise be unused, it would just go to different forms of consumption chosen by the household rather than infrastructure or button museum's or whatever it is that the body politic decides is socially worthy.

You would need to make the argument for directing spending to some project that creates a positive employment externality (which Keynesian economists sometimes do), so that their job produces a multiplier effect that individual household spending would not create. Empirical support for such an externality is very thin, but it can always be argued that it has just not been done correctly in the past and that we will get it right this/next time.

Suzie said...

Then does that mean that instead of considering if the $ 9,000 is worth the creation/saving of 3 million jobs, the question is; "is the output (or completed job) that these 3 million workers are going to complete worth the input of $9,000 to each tax paying individual in comparison to the opportunity costs that the individual may have spent that $9,000 on themselves?"

When thinking this through, I thought that maybe the utility of the jobs may heighten the value of the government projects since the recession has brought our attention to unemployment (even if, financially speaking, their wage is reflective in the overall cost of the project). However, as you pointed out all types of consumption will spur job creation, and while I may value the government project/jobs more than my $9,000, this does not mean every other American does.

Justin M Ross said...

Yes, that would be the correct question, and it would only be worth asking if the directed spending can generate an additional externality that the privately determined spending could not. Also there is the issue of whether or not the spending somehow gets to the "structurally" unemployed, which means we are not just moving employees from one job to another but instead creating a job for someone who was otherwise unable to get one.