Internet penetration is never correlated with increasing power to dictators, and is often correlated with increased freedom

Philip N Howard wonders if there are any countries that have, on balanced, suffered as a result of the coming of the Internet -- say, because improved networks created so many opportunities for dictators to spy on dissidents that it swamped any free speech/free association benefits that the Internet delivered. So he scatter-plotted PolityIV’s democratization scores from 2002/2011, and cross-referenced them with World Bank/ITU data on internet users. The conclusion: by this method, no country experienced a decline in its overall levels of a democracy as it attained widespread Internet penetration, and almost all many countries experienced a rise in democracy levels that correlated to a rise in Internet penetration.

Are there any countries with high internet diffusion rates, where the regime got more authoritarian? The countries that would satisfy this condition should appear in the top left of the graph. Alas, the only candidates that might satisfy these two conditions are Iran, Fiji, and Venezuela. Over the last decade, the regimes governing these countries have become dramatically more authoritarian. Unfortunately for this claim, their technology diffusion rates are not particularly high.

This was a quick sketch, and much more could be done with this data. Some researchers don’t like the PolityIV scores, and there are plenty of reasons to dislike the internet user numbers. Missing data could be imputed, and there may be more meaningful ways to compare over time. Some countries may have moved in one direction and then changed course, all within the last decade. Some only moved one or two points, and really just became slightly more or less democratic. But I’ve done that work too, without finding the cases Morozov wishes he had.

There are concerning stories of censorship and surveillance coming from many countries. Have the stories added up to dramatic authoritarian tendencies, or do they cancel out the benefits of having more and more civic engagement over digital media? Fancier graphic design might help bring home the punchline. There are still no good examples of countries with rapidly growing internet populations and increasingly authoritarian governments.

Are There Countries Whose Situations Worsened with the Arrival of the Internet?


  1. Doesn’t that imply that increased power to dictators is not correlated with increased freedom ? that’s something to ponder here.

  2.  From just the what’s on the graph, we could just as well argue that technology prevents  increased freedom (there are no points in the upper right).

    The headline is somewhere between “flat out wrong” and “misuses ‘correlation’ so badly that it isn’t a coherent argument that can be evaluated.

    And *again* some BoingBoing contributors post things that laugh in the face of other cotnributors’ attemps to support scientific and mathematical literacy.

  3. YEAH YEAH YEAH . I’ve been telling this to people for years – that the Internet empowers people more than organizations – and met with such skepticism! It is so obvious, though, that so many activist groups that have gotten such a supercharge out of the Internet and allied technologies. I’m so glad to have data, primordial as it has. It may be a cliche to say we are entering a second renaissance, but that doesn’t make it less true.

  4. “…by this method, no country experienced a decline in its overall levels of a democracy as it attained widespread Internet penetration…”
    Am I misreading the graph? Because there are definitely points to the left of the central axis, indicating countries with increasing internet access and decreasing democracy. Not as many as on the right, but certainly not “none”.

  5. First, horrible graph… -10 on the Y-axis? Negative penetration?

    Second, as @boingboing-4ce0c21386a1e566e12f73d9c0cff630:disqus said, the analysis isn’t sound. The data shows absolutely no correlation between the variables.

    More importantly, the graph shows rate of change in both variables, not the absolute numbers themselves. We can do better on Gapminder; try running the animation with “Play” if you want to see the changes.

    Look at the top of the graph – that’s where high penetration is. Notice Saudi Arabia, the UAE and Qatar in that top left – that’s where you have high penetration and a low Democracy Index. Yes, there are fewer countries high up and to the left, but at best it seems to be a weak correlation. Highlight countries like Russia, and you’ll see penetration go up and DI come down. In general however, most countries go straight up – increased penetration, no change in DI.

    Cory, I do desperately want this to be true, but unfortunately, it’s not!

  6. The problem already starts with the definition of ‘democracy’. On the PolityIV graphs I see the US counts as a ‘full’ democracy and France as a ‘democracy’. Given that the US has only 2 functional political parties; one more than North-Korea and the bare minimum for any country wanting to pretend being a democracy I would question whatever metric is being applied here.

    A country is not democratic because if you have the correct photo-ID you can vote on a leaky and proprietary computersystem for on of two political parties both financed by unlimited ‘free speech’ donations from Wallstreet and other corporate interests.

    And while living under a monarch can for sure be terrible living in a country that declares itself a democracy while its government has the legal right to kill anyone anywhere for any reason without any judicial oversight including its own (underage) citizens by drone strike does not feel like ‘democracy’ to me.

    So we’re measuring internet acces againt a metric called ‘democracy’ but it is rather unclear what ‘democracy’ means in practice if the US of today is more ‘democratic’ than France.

  7. Comment by David Golombia in the source post. Great point:

    “the PolityIV data speaks solely to the form of government, but not to corporate power.I take a serious consequence of Morozov’s arguments (and my own) to be that corporations often embody authoritarian forms of power, and that digital technology concentrates and reinforces their power greatly. Given the neoliberal emphasis on “democracy” = “free markets” = “corporate freedom,” I would be very surprised not to also see a serious growth of corporate power (both economic and other forms of power), which for many count as a form of democratization, although on some theories of democracy that’s much less true.”

Comments are closed.