A regime is “A set of formal and informal political institutions that defines a type of government.” (Orvis, 2017, Ch 3. Regime change is the process through which one regime is transformed into another (Orvis, 2017, Ch. 9). An institution is a way of organizing social behaviour, which can involve formal rules and organizations, and informal understandings and assumptions about how things are done and who is important. A “democratic” regime is a shorthand expression for a lot of ideas about how a state functions. Similarly an “authoritarian” regime conceals a lot of variables. In the table below, six different regime types are illustrated for 11 countries. Some have had multiple regimes over the last 115 years. Others, the Orvis and Drogus (2017) claim, have continued with a single regime.
What are the specific institutions that define the states as having one or another regime type? What do India and Germany have in common? How are China and Russia similar but not quite the same?
One of the clues is the Freedom House score in the last column – free, partly free, or not free. What is Freedom House, and how do they calculate this score? We can explore the web site for Freedom House. Here are some things you want to know before you use their data: who pays for it? Where is it housed? How credible are its board members and key influencers? How are the indices assembled, i.e. what are their component parts? (You’ll have to look in the methods section of the Freedom House data set). Do these components of the final score fit neatly in the ladder of abstraction below the ultimate “freedom” score? How would you order them? Do you see any issues with multicollinearity, i.e. do you think any of the components are strongly related? How might you check?
Do you think that all the variables related to “freedom” are captured in the Freedom House indicators? Or is it possible that there are unobserved causes of “free regimes” or “liberal democratic regimes” that might be related to other observable phenomena that are included in the index? This would be a problem of endogeneity bias, which is well explained in this short podcast (Endogeneity, 5 minutes) . While science experiments can usually control for endogeneity, true natural experiments or randomized natural experiments such as the ones we rely on in comparative poitics may involve greater risk of endogeneity (video, 4 mins)
Now let’s take a few minutes to think about what is not included in these concepts of regime. We seem to focus on civil, political, and social rights, and protections for the individual. Should we also consider intergenerational rights, collective rights, and individual responsibilities to the state and fellow citizens? What might an index of social responsibility look like? Would this suggest different regime types? If we conceived a different ladder of abstraction, with “sustainability” as the top rung, what might the supporting concepts and resulting indices look like? What about a top rung that was “fairness” or “equity”? How might that index be constructed?
Now do a quick thought experiment: you can see a large field full of “ladders of abstraction” with the top rungs variously labelled: wealth, security, freedom, equity, sustainability, solidarity, justice, happiness, and so on. Policy makers have to decide which ones to climb. How do societies and individuals ascribe weight or importance to these various concepts? What role does measurement play in making these concepts concrete policy objectives?
Now let’s consider the individual regime changes implied by the table above, from Orvis and Drogus (2017) Ch. 9. Considering Brazil, for example, we can see that it democratized twice, and suffered two military coups, one producing a “neofascist” regime in the 1930s, and one producing a “modernizing authoritarian” regime in the 1960s.
There’s no hint in the Freedom House label “free” about whether the “democracies” are solidly democratic, or just barely in the camp. But we can explore this. We can go back into the data and see how much we would have change the component indicators in order for the label “free” to change. How “sensitive” is the indicator to variation? And is the measurement device robust enough for us to be confident that when it does change, it represents are real phenomenon, not a marginal subjective change? To explore this, download the excel file of the Freedom House Data, and play with the numbers to see how big a change is needed in different categories to change the top level of abstraction. Which categories make the biggest change? What are some of the changes that might happen in a “free” country for it to slide into the “partly free” category? What about a “partly free” country like Mexico or Nigeria or Singapore—what might change that would shift it up to the “free” category?
Boix and Stokes, Ch. 14
Boix and Stokes, Ch. 16
Orvis and Drogus, Ch. 9 (Moodle)
Read through this page and consider the questions in the introduction and questions for discussion.
Class coordinator(s) will direct activities over the three periods of the week.
Quittkat, C., & Finke, B. (2008). The EU Commission consultation regime.
Freedom House Indicators methodology https://freedomhouse.org/report/methodology-freedom-world-2018
Polity III as an alternative https://www.icpsr.umich.edu/icpsrweb/ICPSR/studies/6695
How do their methods differ?