Rulers, Elections, and Irregular Governance (REIGN) Showcase: Part I

Democracies versus Non-Democracies
Iraqi police show of their ink-stained index fingers – proof that they visited the polls to cast their ballot in Iraq’s historic parliamentary elections. photo credit: Staff Sgt. Jim Goodwin, USMC

In 2016, Curtis Bell and the One Earth Future Research (OEFR) team launched the Rulers, Election, and Irregular Governance (REIGN) Dataset. These data provide a unique opportunity to examine patterns in regimes over time at the nation-state level. REIGN seeks to provide up-to-date information about the political environment of countries each month. The foundation of REIGN rests on two unique contributions to the measurement of political governance. First, it measures individual and system factors concerning the head executive within a country. Second, it provides systematic information on all elections concerning the head executive. The novel contribution of REIGN in this regard is its measurement of election-specific factors that capture information about election announcements, delays, and outcomes. Alongside this information, irregular leadership change is also captured in the form of violent events and coups.

Recently, REIGN has been used to illustrate trends in the distribution of democratic governance over time. Jay Ulfelder at KOTO used REIGN side by side with Freedom House data on political and civil freedoms to compare time-series trends in within-country freedom with REIGN’s broad regime classification types. Clay Fuller at AEI weighed in on the democracy debate by using REIGN’s regime data to show that democratic systems have steadily increased in number through time and that the majority of the world’s population now lives under a democratic system. Internally, OEFR has also produced a series of analyses concerning autocratic elections, leadership survival, and coups by leveraging REIGN’s novel leadership and election data.

Selected Trends Across Time and Space

Instead of adding to the above analyses, and more specifically the debate concerning the decline of democracy, this post seeks to provide a technical showcase of what REIGN has to offer for both researchers and empirically driven analysts and policymakers. This series will include three posts that unpack selected aspects of the REIGN data to highlight (1) broad regime/non-regime trends across time and space, (2) election/non-election patterns across time, and (3) leader-specific trends at both the individual and regime levels of analysis. This post breaks down the latter point.

As of May 2018, the data contain 130,907 leader-month observations from January 1950 to the present and contain thirty-eight variables. Regarding leadership, REIGN contains information on 2,402 unique leader-tenure periods across 201 countries. In terms of elections, it contains information on 2,946 unique election events across 189 countries.


Democracies Vs. Non-Democracies

Distribution of Democratic and Nondemocratic Regimes Through Time and Space (1950–2018)
Source: Curtis Bell, Rulers, Elections, and Irregular Governance (REIGN) Dataset (Broomfield, CO: OEF Research, 2016). Available at

Because REIGN contains both time-series and geographic information about regimes, leadership, and elections, it is possible to visualize interesting trends in the distribution of regime types across the planet. Above, we visualize the geographic and temporal distribution of democratic (purple) and nondemocratic (gray) regime types.

Cautiously corroborating observed advancement in the adoption of democratic systems, the above visualization shows a steady increase in the global distribution of democratic regimes through time. From a spatial perspective, one can also make some initial observations. As of 2018, democratic systems seem to have made large inroads in both South America and Southeast Asia and moderate gains in sub-Saharan Africa. While these patterns give cause for optimism about the future of democratic governance, they cannot capture variation in more nuanced aspects of subnational freedom and civil liberty. For a more direct comparison of regime and subnational indicators of freedom, see the aforementioned analysis conducted by Jay Ulfelder.

Moving beyond trends in democracy and nondemocracy, REIGN offers insight into non-regimes in the form of warlordism (orange) and foreign occupation (purple). Here, warlordism is defined as the breakdown of functional governance due to conflict. In contrast, foreign occupation occurs when foreign politicians or militaries hold de facto power over governance. The GIF below displays this distribution and demonstrates the sporadic nature of both the breakdown of governance and occupation by foreign forces. Over the sixty-nine years covered in REIGN, eight countries (Bosnia, Liberia, Chad, Uganda, Libya, Yemen, Afghanistan, and Laos) experienced warlordism and eleven countries (Dominican Republic, Grenada, Austria, Bosnia, Kosovo, Cyprus, Iraq, Syria, Lebanon, Afghanistan, and Japan) experienced foreign occupation. One case that comes close but does not quite fit these descriptions is Somalia, which has undergone periods of foreign force involvement but never to the extent that foreign forces held de facto power over the government. Furthermore, Somalia does not meet REIGN’s criteria for warlordism because it has not had a single qualifying leader during its fall into anarchy and total state collapse.

Warlordism and Occupied States

Distribution of Warlordism and Foreign Occupation Through Time and Space (1950–2018)
Source: Curtis Bell, Rulers, Elections, and Irregular Governance (REIGN) Dataset (Broomfield, CO: OEF Research, 2016). Available at


Warlordism occurred in thirty-six of the sixty-nine years with clusters around the 1992–1996 period (163 collective country-months) and the 2012–2017 period (112 collective country-months) with the two periods’ representing 53 percent of all instances of warlordism in the data. Across countries, Laos (161 months), Libya (87 months), and Liberia (83 months) experienced the most time under warlordism and collectively experienced 64 percent of all monthly warlordism periods.

In terms of foreign occupation, sixty-seven of the sixty-nine years had at least one occupied country. Occupations cluster heavily in the 2000–2016 period (742 collective country-months or 41 percent of all occupations) and the 1996–1999 period (147 collective country-months or 8 percent). Taken together, late 1996 onward accounts for almost half of the occupation periods captured in REIGN. Breaking down time trends by country, Cyprus (703 months) experienced the longest foreign occupation and accounted for 39 percent of all occupation months globally. Moving to the next four countries, Lebanon (348 months), Bosnia (301 months), Afghanistan (154 months), and Iraq (108 months) collectively experienced 51 percent of all occupation months captured by the data.

So, why is disaggregation of governance type important? Moving beyond the democracy versus nondemocracy debate, scholars have pointed to the mechanisms for regime emergence, durability, and institutional outcomes. While significant inroads have been made for democracy, particularly in the Western Hemisphere, foreign occupation and warlordism are both on the rise heading into the twenty-first century. Empirical research would do well to monitor trends over time as well as the geographical distribution of non-regime types such as these.

Overall, REIGN is a valuable resource for understanding and visualizing variance in regime type through time and across countries. While the data have been used to understand multifaceted variation in democracy versus nondemocracy, this descriptive analysis demonstrates that REIGN is capable of moving beyond this kind of comparison. Not only does REIGN provide a nuanced set of regime typologies, it also captures non-regimes such as warlordism, foreign occupation, and provisional regimes. The implications of these data are that parts of the world still experience highly irregular systems of governance and that one can utilize REIGN to more robustly analyze patterns in both regime distribution and non-regime distribution.