Internet has limited impact on elections, study finds

A new study finds that the introduction of broadband internet in Germany slightly reduced voter turnout, had little impact on local elections, and did not work to the advantage of smaller parties

The rollout of broadband internet access in Germany does not appear to have had much of an effect on elections results in the country. Nor does it appear to have benefited fringe parties, despite the wider reach it afforded their message. On the contrary, it even reduced their vote count. The real-life effects of new technologies can indeed be counterintuitive.

These are some of the findings of a study based on rich data on local, state and national elections in Germany, combined with unique telecommunications data that document the availability of broadband internet access across roughly 12,000 German municipalities. The data makes it possible to compare voting behaviour before the internet era to that observed after the web started to be used more widely in the country in 2004.

Technological peculiarities of the German DSL network, which basically rests on the voice-telephony network rolled out long before the internet era, provided useful instruments to test the robustness of the findings, in particular as regards causality. One such peculiarity is a technological mistake made early in the efforts to bring East Germany up to par in telecommunications.

The roughly 11 per cent of households in the former East that had no telephone lines were equipped with a so-called OPAL system that used fibre wires, in the expectation that this would become the connectivity standard. It turned out not to be so, with DSL broadband – which is incompatible with the OPAL wires – becoming the standard instead. This made it possible to compare voting behaviour in areas with and without broadband access.

The data shows that an increase in broadband availability from zero to 100 per cent decreases voter turnout by 1.9 to 2.5 percentage points. Given an average turnout of 64.4 per cent, this translates into a reduction of 3.0 to 3.9 per cent. This magnitude is comparable to the effect of the introduction of television in the United States in the 1940s and 1950s, which according to studies conducted in that country reduced voter turnout in congressional races by 2 per cent.

The internet effect appears not to affect local elections, however. The evidence shows that this is due to the fact that the internet tends to crowd out national papers, but not local ones, which may affect the dissemination of supra-regional information, a powerful driver for voting interest in supra-regional elections.

Single parties do not systematically benefit from the increased channels for publicity that the internet offers. Intriguingly, fringe parties appear even to lose voters as a result of the internet. This may be due to the fact that, before the internet, voting for a small party with nary a chance to win a significant vote share was one way to make a political statement. Nowadays, blogging and other social media may offer a more effective way to express political opinions on specific topics. This, together with the effect of social media, would make a fine subject for further study.

Thus, the initial introduction of the internet in Germany did not shake up election behaviour all that much. But what has happened afterwards is another story. Would the German Pirate Party, for instance, even exist without the internet? The odds are it would not.

Oliver Falck is from the Ifo Institute in Munich. With Robert Gold and Stephan Heblich he is the author of E-Lections: Voting Behaviour and the Internet