Spitzer on the Stimulus Plan

Eliot Spitzer, in an apparent bid for a political comeback, has taken to writing op-ed pieces for Slate. A couple of these in the past have been pretty good, but today's piece, Robots, not Roads, was a mountain of bad advice.

Spitzer criticized Obama's stimulus plan, which is largely a combination of tax cuts and infrastructure spending, as being too backward looking:

Paving roads, repairing bridges that need refurbishing, and accelerating existing projects are all good and necessary, but not transformative. These projects by and large are building or patching the same economy with the same flaws that got us where we are. Our concern should be that as we look for the next great infrastructure project to transform our economy, we might rebuild the Erie Canal and find ourselves a century behind technologically.

Spitzer instead suggests smart power meters, which would adjust the cost of power consumption based on usage and other factors, alternative fuel infrastructure (hydrogen pumps, recharging stations, etc.) and giving every school a robotics club. (Spitzer does suggest using some of the stimulus for subsidizing the switch to online medical records, which I totally support).

There are three main problems with Spitzer's high-tech stimulus plan. The first is that for the bulk of this spending (other than the robotics clubs, since giving every one of America's 120,000 schools 10 grand for a robotics club would only cost a bit over a billion), there isn't a readily agreed upon standard that can be quickly implemented. Electricity transmission, medical records and auto fuel all are governed by network effects, whereby the adoption of a standard makes each unit much more effective. A stimulus package that gives Georgia money to implement one kind of medical record program isn't very helpful if New York is implementing a different one- which is what would happen without clear standards. Same thing with alt fuel- even Spitzer isn't clear whether electric vehicle recharging or hydrogen is the way to go. Spending the stimulus that way would be like paying for a mix of canals, highways and railroads that aren't interoperable.

The second problem (related to the first) is that none of this stuff is quite ready to go yet. The benefit of spending on roads, buildings, etc. is that most states and municipalities have plans and projects that could start hiring immediately once the federal checks show up (or even once they're promised). Many of these projects, like NYC's 2nd avenue subway line, or construction on I-81 through PA, were already in progress before state funds dried up with the recession. High-tech stimulus would require an enormous amount of time figuring out which projects are worthwhile, how they should be implemented, understanding the environmental impacts, etc. before anyone could break ground. With the economy in need of an immediate jumpstart, this kind of delay is a no-go.

The third problem is that high-tech stimulus would do less to ease unemployment. Many of the recently unemployed are low-skilled workers or workers in the construction trades, who can readily find work in standard infrastructure projects. They then spend their wages on local goods and services. Standard infrastructure projects also use a lot of bulky, inexpensive stuff like concrete, wood, steel, etc., much of which is produced in the U.S. because it's not feasible to ship from far away (steel largely excepted). The cement or lumber companies then pay their employees who spend that money locally. This creates what's called a Keynsian multiplier, which increases the value of every dollar of infrastructure spending. High-tech stimulus, however, would probably require importing (at least initially) many of the components of the smart meters, computers for medical records or technology for recharge or hydrogen stations- and that money would go straight out to the exporting countries. Additionally, although there have been layoffs in the tech world, the U.S. still has a shortage of engineers and high-tech workers, and would probably have to either open up immigration or repurpose engineers who were previously working on other projects in order to implement a massive, multi-hundred billion dollar high-tech stimulus plan. Either way, it wouldn't help much with unemployment.

Finally, one of the best reasons to undertake a standard infrastructure stimulus program is that, because there's little demand for construction labor and materials, those "inputs" would cost less than they would in better times, so the government could buy a lot of infrastructure for each stimulus dollar spent. With American infrastructure slipping into third-world territory, now is a great time to buy.