Has the world taken an authoritarian turn, or is talk of a democratic recession overblown?

Earlier this week, Freedom House released its annual report on freedom in the world (here), and it paints a worrisome picture. For 12 years in row, the group has observed declines in global freedom, and in 2017 “democracy faced its most serious crisis in decades.” As the chart below shows, the average freedom score among countries worldwide fell again last year to its lowest point since Freedom House began reporting this 100-point scalar measure in the early 2000s.


Meanwhile, One Earth Future Foundation (OEF) recently updated its global data on countries’ types of government, REIGN (here), through the end of 2017, and that source suggests a less troubling outlook. The chart below plots change over a much longer period of time in the shares of countries worldwide with democratic or various non-democratic national political regimes, according to REIGN. As the chart shows, democracy became the predominant form of government worldwide in the 1990s, and the global share of democracies has not declined in the past decade.1


So, why do these two well-regarded and carefully constructed datasets tell what seem to be such different stories?

The divergence stems primarily from differences in the nature of the two measures. Following a long scholarly tradition, OEF aims to sort the world’s governments into bins representing different patterns of gaining and holding political power. It is less concerned with minor variations within those categories than it is with correctly capturing these grosser differences in type. By contrast, Freedom House’s annual report focuses on the broad concept of, well, freedom, and its dataset aims to observe incremental gains and losses on that single scale. In other words, where Freedom House tries to capture differences of degree, OEF focuses on differences of kind, and changes visible through one of those lenses won’t always be apparent through the other.

In this view, we see clear declines in freedom over the past decade among democracies and, especially, personalist autocracies.

The importance of this difference becomes clearer if we look at trends in Freedom House’s freedom scores within our aggregations of REIGN’s categories. The density plots below show changes in the distribution of Freedom House's freedom scores from 2007 to 2017 for countries within the various REIGN regime types.1 In this view, we see clear declines in freedom over the past decade among democracies and, especially, personalist autocracies.

  • The center of gravity for personalist autocracies, the largest set of non-democracies in the post–Cold War period, has shifted a solid 15 points to the left. Already the least free type of regime on average, they appear to have grown even harsher in the 2010s. Illustrative examples include Russia, Azerbaijan, Kazakhstan, and the Democratic Republic of Congo.
  • Among democracies, the distribution of freedom scores has also slumped a bit toward the more repressive side of the scale, as countries like Poland, Hungary, and the United States have seen their scores decline without crossing the threshold to full-blown authoritarianism (per REIGN).
  • By contrast, hybrid autocracies appear to have become a bit freer on average than they were 10 years ago, but there are also fewer of them (see the REIGN chart above).
  • The global share of one-party autocracies—places like China and Vietnam—hasn’t changed much in the past decade, nor has the distribution of freedom scores among them.


So, it turns out, both summary statements can be true. The share of governments around the world that are effectively democratic has not changed much in the past 10–15 years, but many countries on both sides of the democracy/autocracy divide have suffered noticeable declines in freedom.

1. The REIGN data sort regimes into 14 categories. We have aggregated some of those categories to make them easier to compare. In our chart:

  • Parliamentary and presidential democracies both go in ‘democracy’;
  • Military, indirect military, and provisional military regimes are all labeled ‘military autocracy’;
  • Personal dictatorship is relabeled ‘personalist autocracy’;
  • Oligarchies and dominant party regimes comprise ‘one-party autocracy’;
  • Military-personal, party-personal, party-military, and party-personal-military hybrid regimes go into ‘hybrid autocracy’; and
  • Monarchy, provisional civilian, foreign/occupied, and warlordism get lumped into ‘other’.

For brief descriptions of each regime type, see REIGN’s documentation. For key scholarly works discussing these concepts, see here and here.

2. Military autocracies are omitted from this chart because there are too few of them in 2017 to generate an informative density plot.

Jay Ulfelder

Political science Ph.D. (Stanford 1997), research director for the US government-funded Political Instability Task Force (2001-2011), Good Judgment Project superforcaster.