As Thomas Friedman’s latest brainwave reminds us, ‘four more years of the same’ isn’t just a fading echo from the Romney campaign; these days it’s also a handy approach to writing for the New York Times. While most of the political world is focused on the fiscal cliff, Friedman is focused on another round of fiscal stimulus. He’s quite taken with the amount of money we could put towards ultra high-speed internet networks like the one in Chattanooga, Tennessee, though he’s unwilling to admit the true cost, and intent on distorting our sense of the benefits. He concludes his column with a proposition:
[Chattanooga's] network was fully completed thanks to $111 million in stimulus money. Imagine that we get a grand bargain in Washington that also includes a stimulus of just $20 billion to bring the 200 biggest urban areas in America up to Chattanooga’s standard. You’d see a “melt-up” in the U.S. economy.
First, the math. Friedman uses Chattanooga’s stimulus grant amount as a proxy for the average cost of extending the network to each of 200 cities. This is a bogus assumption, meant to downplay Friedman’s expensive ambitions. We don’t have to scour municipal accounts; just read the rest of Friedman’s column. A few paragraphs earlier, there’s this:
[Chattanooga passed] a $229 million bond issue to build a world-class fiber-optic grid
In yet a different paragraph, Friedman suggests the bond issue occurred in 1997. (no word on maturity or interest rate)
In summary, the $111 M stimulus grant only covered the cost of finishing the network (making it “fully completed”, to quote Friedman). Commencing the project cost no less than $229 M. Total project cost: $340 M. Using this more accurate number as the average cost for the top 200 cities, the cost of Friedman’s project adjusts to $70B – several times larger than the $20B he claims.
Then there’s Friedman’s ‘fallacy of composition’, which amounts to an even greater deception. In short, even if the benefits for Chattanooga exceed $340 million, they aren’t widely replicable. Indeed, attempts to replicate them in other cities could undermine the success of Chattanooga’s project. The greatest benefit for Chattanooga is in becoming a destination for companies drawn from around the country, or for start-ups that would have been started elsewhere; essentially a shifting of jobs to Chattanooga from the places where they would exist otherwise. If similar networks are built in other cities, this ‘destination effect’ becomes weaker. If enough are built – even a fraction of the 200 Friedman wants – the effect disappears completely.
If Friedman had cited Las Vegas for why every state should legalize gambling and build high end hotel-casinos, the fallacy would be obvious. It’s less obvious here because, unlike having a casino within easy driving distance, having higher speed internet access does actually tend to improve productivity of the existing workforce, creating at least a few jobs that would not exist otherwise anywhere in the system. It’s easy to conflate this very modest productivity effect, which actually is widely replicable, with the substantially larger ‘destination effect’, which lasts only so long as your city is both the first, and the only mover. Buying Freidman’s pitch requires conflating these distinct types of benefits, or simply assuming that they account for roughly equal parts of the aggregate benefits experienced in Chattanooga.
In short, Friedman is taking his readers, and American tax payers, for dupes. Truth is, most of the benefit comes from the destination effect, with only a small fraction coming from the tiny upticks in productivity associated with a faster internet connection. We learned this when, in 2011, the left ‘debunked’ the Texas miracle under the economic stewardship of Rick Perry. The left explained how low tax rates and measured regulation only makes sense as way to steal businesses from Democratic states. As national policy, it just won’t work–they assured us. Yet, Friedman wants us to follow the lead of . . . the Mayor of Chattanooga? Of course, Chattanooga is located in a conservative state, one with levels of taxing and regulation comparable to Texas, one presumes.
Adjusting for the fallacy of composition and off-budget accounting, which we do by assuming that 10 percent of the total benefit comes from productivity gains and multiplying the $111 million stimulus price tag by three to mimic the Chattanooga project’s true cost, we see that Friedman implicitly overstates the expected return on investment (ROI) by a factor of 30! And that assuming the true ROI is positive. Although more optimistic assumptions on productivity gain would make Friedman’s exaggerations appear slightly smaller, it’s more likely that his ‘new Neal Deal’ is net negative proposition for most every city – say, 190 of the top 200 – and thus a huge net negative for the country as a whole.
And this is before you adjust ROI for interest payments, since a new stimulus would be funded with borrowed money. Even if the promise to repay is a lie, and therefore the cost of borrowing illusory, the ‘investment’ loses nearly all its potential benefit once it moves from anecdote to nationwide model. On that level, it becomes a make-work internet arms race where the whole nation comes out behind – perhaps Chattanooga included, depending on how long their internet advantage lasts. (It should last a while, considering the 15-year lead time.)
The productivity gains, such as they are, may increase inequality. This has been the general experience with the spread of information technology over the past twenty years. In theory, these technologies increased productivity across society. In practice, certain people increased their productivity more than others, while the managerial and executive classes were best able to leverage these gains for personal benefit. Continued marginal advances in technology should bring more of the same. And since the aggregate effect falls well short of what Friedman’s fallacy suggests, universal high-speed internet is unlikely to produce enough new jobs to justify increasing inequality.
And then there’s the rhetorical questions: didn’t we have a $800 billion stimulus in 2009? And if this idea is so great, how come it accounted for only $111 million of those expenditures? Perhaps the entire $70 billion tab couldn’t be met, but a mere 0.15 percent suggests low confidence. And are there other projects from the stimulus that deserve a 700 fold funding increase during Obama’s second term? One can imagine the numbers adding up quick.
Most importantly, is more stimulus the Democrats’ big new idea? A Times column is the place for fresh thinking, should such a thing exist on the left. If Friedman is any guide, Obama’s updated approach is to super-size the underwhelming policies of the last four years. But as the fallacy of composition counsels, sometimes less is more.