Latest Empirical Findings on Democratic Effects of the Internet

Jacob Groshek from Iowa State University recently published the latest results from his research on the democratic effects of the Internet in the International Journal of Communication. A copy of Groshek’s study is available here (PDF).

Groshek published an earlier study in 2009 which I blogged about here. In this latest set of findings, Groshek concludes that “Internet diffusion was not a specific causal mechanism of national-level democratic growth during the timeframe analyzed,” which was 1994-2003. The author therefore argues that “the diffusion of the Internet should not be considered a democratic panacea, but rather a component of contemporary democratization processes.” Interestingly, these conclusions seem to contradict his findings from 2009.

The purpose of this blog post is to summarize Groshek’s research so I can include it in my dissertation’s literature review. What follows therefore are some excerpts that summarize Groshek’s research design and methodology. I also add my thoughts on the study and the implications of the findings.

Some Background:

“Technological developments, especially communicative ones, have long been positioned — and even romanticized — as powerful instruments of democracy (Dunham, 1938; Lerner, 1958). This tradition goes back at least as far as the printing press and its contribution to democratic movements of past centuries (Schudson, 1999) in relation to conceptions of the public sphere and the fourth estate (Jones, 2000). Over the course of the past century, telegraphs, telephones, radios, and televisions were all introduced as “new” media, and each of these technologies were often ascribed broad potential for enhancing democratic development around the world (Becker, 2001; Navia & Zweifel, 2006; Spinelli, 1996).”

The Methodology:

“Though there are many ways to operationalize democracy and measure the prevalence of media technologies, this study relies principally on macro-level time–series democracy data from an historical sample that includes 72 countries, reaching back as far as 1946 in some cases, but at least from 1954 to 2003. From this sample, a sequence of ARIMA (autoregressive integrated moving average) time–series regressions were modeled for each country for at least 40 years prior to 1994.”

“These models were then used to generate statistically-forecasted democracy values for each country, in each year from 1994 to 2003. A 95% confidence interval with an upper and lower democracy score was then constructed around each of the forecasted values using dynamic mean squared errors. The actual democracy scores of each country for each year from 1994 to 2003 were then compared to the upper and lower values of the confidence interval.”

The Results:

“Based on the statistical findings, three countries that demonstrated democracy levels greater than those statistically predicted  [Croatia, Indonesia and Mexico] were selected for brief contemporary historical analyses to identify whether the Internet acted as a specific causal mechanism that may have contributed to democratization processes. These case study evaluations were basic overviews of historical events, figures, and policies that placed these findings into context to better specify what precise role, if any, the Internet had on the increases in democracy observed in these three countries that were greater than they had been predicted to be, statistically.”

Interestingly, out of the 72 countries studied, the only one with democracy scores significantly below the statistically predicted score was Belarus.

“While the purpose of this study is to more specifically assess the possibility that Internet diffusion might be linked to democratic growth, the case of Belarus provides an important counterbalance to that concept. This is because, starting with 1995, the actual democracy score was less than the predicted democracy score — and it remained below the predicted values through 2003, even though Internet diffusion reached approximately 14% by the end of the time frame investigated. Thus, it is evident that less democratic countries can invest in increasing Internet diffusion and still constrict democratic development.”

What about Croatia, Indonesia and Mexico?

“A circumspect approach to understanding the role Internet diffusion played in Croatia’s democratization is to recognize that, by most accounts, it was an important factor that helped determine the trajectory of political development in this country. It was not, however, the defining feature of this democratic transition, which was set in motion years earlier by a coalescing of events and political figures that also transcended Croatia’s national boundaries (Hampton, 2007).”

“Indonesia had observed actual democracy levels greater than that of the predicted confidence interval from 1999 to 2003. Yet, for nearly all of the timeframe investigated here, Indonesian media development was tightly restricted by the government and subject to severe censorship (Eick, 2007), so it seems unlikely that the diffusion of the Internet would be a critical democratic agent. In addition, the diffusion of the Internet was a paltry 0.44 people per 100 in 1999, when the democracy level spiked through the upper confidence interval of the predicted value.”

“[In the final case, it is] impossible to summarily conclude that Mexico was more democratic precisely due to Internet diffusion than it would have been had the Internet not diffused, at least when considering institutionalized national level democracy. This is because the transnational civil society network pioneered by the Zapatistas was more about élites who had Internet access and how the Zapatistas tapped this group and projected their ideological views through the Internet, even though, in Mexico, the Internet only reached a tiny portion of the general population. Therefore, it was not high levels of Internet diffusion among the Mexican citizens in 1994, but rather influential Internet users that contributed democratic change during that time period.”

In Conclusion:

“The results of the investigations undertaken in this study yield no conclusive evidence that the democratic growth from 1994 to 2003 was due singularly, or even primarily, to the diffusion of the Internet.”

Side note: I personally don’t know anyone or of any empirical study that claims that democratic growth around the world is singularly or even primarily due to the Internet. Do you?

“It is therefore prudent to consider the Internet a potentially potent but underutilized democratic tool, one that is only as useful as the citizens who employ and implement it for political purposes (Schudson, 2003). Thus far, the Internet has not been diffused or activated to an extent that this technology has sustained the third democratic wave (Huntington, 1991). Importantly, virtuosity and democratic agency are not inherent in media technologies, no matter how interactive or participatory. Rather, these exist in individuals, and in the crucial applications and uses they make of communicative technologies (Nord, 2001; Schudson, 1999, 2003).”

“Thus, the general conclusion of this study is that the Internet has not catalyzed transformative, national-level democratic growth, although there is some reason to believe that it may contribute to these changes, as the cases of Mexico and Croatia exhibit. This finding, of course, does not rule out the possibility that there may be national-level democratic effects related to Internet diffusion in the future, nor does it rule out possible effects on personal or other sub-national levels.”

It’s great to see more data-driven research on this topic and be spared (albeit temporarily) anecdote-laden and chronically repetitive popular media reports on technologies being either all-liberating or all-repressive. A possible corollary to Groshek’s  findings is that the use of the Internet by repressive regimes did not lead to a statistically significant decrease in expected democracy scores.  In other words, dictators may love the web, but that romance ain’t having a macro-level impact on the level of repression.

Obviously, multiple factors contribute to democratic processes and transitions. The more interesting questions, in my opinion, are these: what are the underlying drivers of protest movements and how might new technologies accelerate those drivers and/or create new ones? Along these lines, how do tactics and strategies from civil resistance benefit from using new technologies? Does the careful, planned and innovative use of these technologies in social protests provide activists with a competitive edge they didn’t have in the past?

Update: My colleague Mary Joyce makes an excellent point regarding the time span covered by the analysis, i.e., through to 2003. As she rightly notes, major social media platforms used for activism, like YouTube (2005), Facebook (2004) and Twitter (2006), were created after 2003. See her blog post here for more of her analysis on Groshek’s work.

10 responses to “Latest Empirical Findings on Democratic Effects of the Internet

  1. Pingback: Too Early to Discount Internet’s Democratizing Effects | meta-activism project

  2. I think Groshek’s conclusion’s are pre-mature, given that the period of greatest online political activism is excluded form his data set. I wrote a post on it, referencing the GDADS:

    • That’s a good point re the data/analysis being limited to 2003 and before. By the way, was just going to email you and recommend that we talk to Jakob about GDADS. I was in touch with him last year and he was keen to co-author some research. Want me to do an intro?

  3. Pingback: Impact of Technology on Democracy and Activism: Findings from Multiple Statistical Studies | iRevolution

  4. Hi Patrick,

    Thanks for the blogging about my paper. I don’t consider the findings of this article contradictory to my Gazette piece (i.e., internet contributes most in partially democratic countries where democratization trends are already underway) but I do completely agree that the time frame ended too soon. That was an artifact of having to finish my dissertation, but I’m not in the process of updating the dataset through 2010.

    The results might be much different, but I suspect some similar patterns of conditional influence might emerge. I can’t say for sure, though, and I’m really looking forward to analyzing the new data because so much has happened in both media and, of course, politics since ’03.

    BTW, I left Iowa and I’m now at Erasmus University in Rotterdam, the Netherlands. Definitely still up for some collaborative activities as well–keep in touch.

    • Hi Jacob!

      Many thanks for your comment, I really appreciate it. Would be great to collaborate on future research, I would definitely welcome the opportunity. In terms of the results of your 2009 and 2010 studies, how do they not contradict each other somewhat? Also, when you wrote “but I’m *not* in the process of updating the dataset through 2010″ did you mean “but I’m *now* in the process…”?

      Thanks again for weighing in. Hope NL is treading you well.

      All best,

  5. Patrick,

    Yes, “now in the process of updating.” It is true, I suppose, that the ’09 article is a bit more rosy, but thematically I’d say they both follow a similar line even though I aggregated countries in the ’09 analysis and examined them individually in the ’10 publication. It is interesting to read your interpretation of contradiction but I would frame that as: even though there were some positive, significant relationships between internet and democracy in the ’09 article those were not so strong as to have actually contributed to individual countries achieving a greater level of democracy than could have been expected otherwise (at least statistically). Perhaps that helps clarify my perspective a bit?
    In my opinion, both studies land on the idea that it isn’t the technology alone, but the sociopolitical environments and the uses people make of the technologies within those environments. Of course, there does seem to be a lot of technological potential that has gone largely untapped in both studies, perhaps partly due to the timeframe ending in 2003.
    An important, ongoing process of analysis. I’ve taken a quick look at your dissertation on here and it looks like you are doing some very nice things.
    Rotterdam is working out really well. Hope all is good on your side.

  6. Pingback: 1 | Pearltrees

  7. Pingback: Education | Pearltrees

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s