Broadband Stimulus: the Case for Going Slow

President-elect Barack Obama, who ran as the high-tech candidate, has promised to make federal investment in broadband Internet service part of both his economic stimulus program and his plan for long-term economic growth. The idea has provoked a flood of ideas, from advocacy groups and the struggling tech industry. But Obama should be very cautious in assessing the untested notion that pumping money into broadband can do much to boot the economy in the short term. Indeed, there’s a good chance the new Administration may end up disappointing its enthusiastic tech supporters. While some parts of its economic program are well fleshed-out, technology policy is little more than a blank page. The concept of broadband stimulus consists of a single vague sentence in the Obama economic agenda.

There's a good reason for the vagueness. I have yet to see anyone make a compelling argument for how spending on broadband might stimulate the economy in the short term. Supporters mostly cite a published by the Information Technology & Innovation Foundation that argues that $10 billion of federal spending for broadband would produce 498,000 jobs. Unfortunately, the evidence presented in support of this rather improbable claim is gossamer thin; no government program in history has come close to generating that much employment at that low a cost, barely $20,000 per job. In purely economic terms, its arguments simply ignore the question of the law of diminishing returns to scale , which means that the efficiency of investment declines as you invest more. And while the ITIF is non-partisan, the group is backed by industry forces that have a lot to gain from such a program; its board includes representatives of Cisco, IBM, Oracle, and Microsoft.

Furthermore, while there are many reasons that country could benefit from faster Internet connections, there's no evidence of a broadband crisis that the government must step in to fix. It's hardly a point of pride that the U.S. ranks 15th out of 30 countries in the Organization for Economic Cooperation & Development's June, 2008 of the proportion of the population with access to residential broadband. But it is worth noting that the countries ranking significantly ahead of the U.S. are nearly all small and very densely populated. The fact that Denmark is 12 points ahead of the U.S. on broadband penetration is not a cause for panic. Of the countries ahead of the U.S., only Korea and. to a limited extent, Canada and Germany, are serious competitors in global tech markets. Over the longer term, improving the nation's broadband infrastructure may boost growth and improve competitiveness, though it is hardly a magic bullet. There are three big issues that need addressing. First we have to decide just what is required. The OECD defines broadband service as anything faster than 256 kilobits per second, ludicrously slow in an age of Internet video. But the argument of advocacy such as that the government should aim for 100 megabit connections, strikes me as expensive overkill, even in the name of “future proofing.” I love my 20 megabit Verizon FiOS connection, but I have never come close to using half its speed and I'm not likely to in the foreseeable future. The 5 megabits or so offered by cable and fast DSL seems a much more affordable goal. The most pressing problems are bringing broadband to those who can't get it and those who can't afford it. The former are 10 million or so mostly rural households and the cost of serving them will rise sharply as we get down to pockets of really isolated folks. the best source of funding for this is repurposing of the obsolete federal Universal Service Fund, which exists to subsidize rural voice lines. The other consists largely of is the urban poor and elderly. The technology divide that separates them from the digital world is unlikely to be bridged just by offering cheap broadband, Subsidies for network construction and subscriptions could end up doing little more than boosting the semi-monopolies enjoyed by the cable and telephone carriers. The digital divide is an important social issue, but one whose resolution requires more than simply throwing money at the problem.

Building a broadband plan for long-term growth is going to take more time than the Administration and Congress can afford to a stimulus package. For sensible policy, slower may be better.

Before it's here, it's on the Bloomberg Terminal.