Inequality Regimes and Rawlsian Growth Rates: Some thoughts on the evolution of inequality 1960-2010, with special reference to Venezuela

(Mostly an excuse to play with the Standardized Worldwide Income Inequality Database, compiled by Frederick Solt, which I just discovered. This post also belongs, loosely speaking, to my long-running series on the quantitative history of political regimes. R code for everything in this blogpost available in this Git repository; you would need to download the dataset separately).

Inequality is difficult to measure. Socially relevant inequalities are manifold, and measurable inequalities in money income are not always especially important. (In the formerly communist states of Eastern Europe income was very evenly distributed; yet this did not mean that there were no important social inequalities). Even inequality in money income is not easy to measure properly. Most existing data is not very comparable accross countries or years, and it is often not even clear to what income concept the sorts of inequality measures people typically use to make a point in political discourse refer to: does it refer to after-tax, after-transfer income or to "market" income? Does it refer to individual or household income? What sorts of things are counted as "income"? How do we account for access to high-quality public services? At best, measures of income inequality are uncertain estimates of an unknown distribution of potential living standards, more or less valid for societies where "money income" is a useful proxy for the ability of people to enjoy various important goods, and of little value outside the context of a conception of a "just distribution" of these capabilities.

Despite the fact that estimates of what is essentially a statistical abstraction often play surprisingly big roles in current political debates (cf. the debate over The Spirit Level some years ago, or recent concern about a rise in New Zealand's gini index), I have recently become more skeptical about the importance of measured income inequality in politics: whatever importance the actual distribution of incomes has for political life in a society, it has to be mediated through complicated processes that refract local lived experience through the prism of context-dependent fairness norms that are only vaguely related, if at all, to the numbers used to measure its skewness.

Yet I'm curious: what sorts of income inequality have in fact increased, and where? How might these changes have mattered? Enter a new and shiny dataset: the Standardized Worldwide Income Inequality Database, which promises to ameliorate some of these measurement problems. The database uses the Luxembourg Income Study - very high quality income inequality data - to calibrate the much larger but less comparable United Nations University World Income Inequality Database. The result: lovely long time series estimates of both the market and the after tax, after transfer (net) gini index of inequality, including standard errors, for 153 countries. My first though on learning about this was: graphs! (Hopefully some non-obvious facts are also involved below).

Let's start by looking at a country that has often been cited as a great success in reducing inequality: Venezuela during the Chavez years. (Also, I was there recently visiting family after a long absence, and I have a personal interest in understanding the changes that have occurred during that time). One question I've been curious about concerns the evolution of inequality in Venezuela relative to other Latin American countries, especially since the coming to power of Chavez in 1999. How do changes in inequality in Venezuela compare to changes elsewhere?

In the following plot, we see changes in both the market ("equivalized (square root scale) household gross (pre-tax, pre-transfer) income", if you must know) and the net gini index of inequality (after tax, after transfer) in 19 Latin American countries from 1999 until 2010, ordered by the estimated rate of inequality reduction (countries that reduced inequality faster appear earlier; read the graph from right to left, top to bottom):

Inequality in Venezuela has indeed decreased relatively quickly since 1999 - the second fastest decrease after Ecuador, which has also had left-leaning governments (though a far more unstable political context, with five different presidents since 1998). Three things are worth noting about the context of these trends, however.

First and most important is that this reduction in inequality is not driven by direct redistribution: there is barely any difference between the "market" gini index (our measure of inequality before taxes and transfers) and the "net" gini index (our measure of inequality after taxes and transfers). To the extent that the reduction in inequality is the result of government action rather than something else, it must have come about through measures like investment in human capital and labor market policies (see Morgan and Kelly 2012, ungated here, for the proper peer-reviewed argument). This is true of all Latin American countries save for Puerto Rico (which is part of the USA in a sense) and (to a lesser extent) Brazil; indeed, redistribution in some countries (Peru) appears to have perversely increased inequality.

Second, most Latin American countries have experienced reductions in inequality during this period, though most remain highly unequal. But Venezuela was already among the most equal countries in Latin America; in 1999, only Uruguay and Costa Rica had lower measured inequality (and the difference in net gini was within the margin of measurement error, so it should probably be disregarded). This surprised me; I had expected higher levels of inequality in Venezuela when compared to other countries, given the level of class conflict on display during the Chavez era. More surprisingly perhaps, if we take a broader look we discover that inequality in Venezuela appears to have been remarkably stable over the past fifty years, fluctuating around a flat trend:

(Lines around dots represent 95% confidence intervals).

In fact, the low level of inequality in Venezuela as of 2010 only returned the country to the level of inequality it last experienced around ... 1992, the year of the February coup which made Chavez famous, and three years after the "neoliberal" paquete of 1989 (which was supposed to have triggered the Caracazo). Inequality did increase after 1992, and poverty had increased before then - the Venezuelan economy had been in decline for a while, as we can see below. Which all goes to show, I suppose, that political unrest and lived experiences of injustice are only very loosely connected, if at all, with measures of income inequality; whereas austerity and large income losses appear more immediately important to political outcomes (as Jay Ulfelder argues here):

(GDP per capita data from the Penn World Table v. 8.0).

Finally, it's probably worth noting that Venezuela's economic fortunes are deeply tied to oil prices, and that the rapid reduction in inequality in the last decade or so should also be placed in the context of the very large rise in the value of oil and gas during this period. Here is an estimate of the per capita value of oil and gas exports for Venezuela, from Michael Ross' oil and gas dataset:

In fact, Venezuela and Ecuador, the countries that have experienced the fastest inequality decreases, have been precisely the two countries that have benefitted the most from oil and gas price increases - money that flows directly to the state (especially since the Chavez government systematically asserted control over the state oil company) and can be used to provide employment and subsidize education, healthcare, housing, staples, and other goods, however inefficiently (e.g., the varios "Misiones" and other social programs created by the Chavez government). At least some of these programs must have played some role in the reduction of inequality, but given the amount of oil and gas money flowing directly to the Venezuelan state (representing most Venezuela's exports, which have become substantially less diversified over the last 15 years) and the typical patterns of clientelism and electoral politics in Venezuela it would have taken a bloody-minded kleptocrat not to reduce inequality by some amount. At any rate, inequality and poverty also diminished quite a bit during the 1970s oil boom, likely through similar channels - massive amounts of money flowing through the state, which increased its ability to employ people and subsidize public services. (I don't mean to sound grudging; though I have doubts about the effectivenes of some of these programs, some of the new housing built during the Chavez years looks decent, for example).

Let's take a broader look, however. How does the Venezuelan experience of inequality reduction compare to some countries outside of Latin America? Just because they've been in the news, let's look at the Venezuelan experience in coparison to Turkey and Egypt; and add the USA and New Zealand to see how two "developed" countries look as well.

That's right: Egypt and Turkey apparently reduced inequality faster than Venezuela in this period (though the error estimates of the gini index for both are also larger), and were less unequal than Venezuela by the end of the period! Also, despite the fact that the degree of "market" inequality was higher in the USA and NZ than in Venezuela, and did not decrease or even increased a little during this period (as measured by the net gini index), both countries remain less unequal than Venezuela (as measured by the net gini index), due to the effectiveness of their redistributive measures.

Now, this is perhaps surprising, but a bit of an aberration, and really, we are dealing with imperfectly estimated quantities (rates of decline) based on measurements with error. So there's really no point in arguing about whether inequality in Venezuela has in fact decreased faster than in Egypt or not; our methods of measuring inequality don't allow us to give a very precise answer to this question (error bars are large, etc.). In any case, it is clear that income inequality has declined pretty fast in Venezuela over the last 15 years, even allowing for some meaurement and estimation error, as we can see by calculating the trend rate of change in the net gini coefficient (the slope of a regression of log(gini_net) on year, to be technical, which yields the estimated trend annual percent rate change in the gini index) for all countries in the dataset:

(I took out countries that had too few datapoints, since the trends didn't look to me like they could be informative. The error estimates in the graph are nevertheless probably too small, since one would need to use the proper rules for error propagation to calculate them, which I have not done. Interestingly, the estimate of the rate of change in the net gini index for New Zealand and the USA since 1998 suggests basically that they have experienced no significant change in measured inequality from 1999 until 2010, contrary to popular belief; their important increases in inequality occurred earlier. More on this in a minute).

What strikes me about this graph is that countries that have achieved very fast reductions in inequality over this period appear to be quite disparate; though left governments are in evidence among these, many countries apparently achieved fast reductions in inequality with supposedly "neoliberal" policies (e.g., Egypt and Turkey) that are now in turmoil. Maybe this is evidence that the gini index does not capture socially relevant changes in inequality; but then it would also fail to capture changes in inequality in non-neoliberal Venezuela and Ecuador. (Of course, other confounding factors may be at work too).

It is also curious that many of the fastest reductions in inequality have occurred in states that do not engage in a lot of explicit redistribution. In fact, a simple correlation between the average redistributive capacity of a state (measured by the percentage difference between the "market" gini and the net gini coefficient) and the rate of decrease in measured "final" inequality over the period is slightly negative (so fastest reductions in inequality have occurred in states that are unable to affect market gini very much at all, or that even increase it through perverse redistribution):

It's probably not worth making too much of this correlation (measurement errors, the relatively short time period under consideration, and confounding factors are not taken into consideration), though it does suggest that many changes in inequality seem beyond the control of most governments. But even when we expand the period of observation all the way to 1960, the correlation does not entirely disappear, though it weakens greatly:

Ultimately, however, the more directly and explicitly redistributive the state has been, the more equal it also appears to be over the long run:

Or, to put it crudely, since 1960 at least market inequality has only been reliably reduced in states that take from the rich and give to the poor. And yet actually taking from the rich and giving to the poor seems to put nontrivial demands on state capacity and political life (witness the existence of robber states that take from the poor and give to the rich). The degree of change in the market distribution of income even appears to be a fair measure of that capacity; from the graph above, it's likely that a state that can consistently reduce gini index of market inequality by at least 30% is a pretty "strong" state (in the "infrastructural" sense of strong), whereas a state that cannot make a dent on the market distribution of income is more likely to be "weak" (with some communist exceptions like the USSR that did not engage in a great deal of explicit redistribution, since, to put it crudely the state owned everything and everyone more or less got paid the same).

And abilities to redistribute income appear to be remarkably "sticky." Few countries appear to become more able to affect the gini distribution over time:

(I've deleted cases with very few data points to make the graph look prettier. See the code for the details). What is striking about this graph is how stable the redistributive capacity of most states has remained over a period of more than six decades: many countries show basically zero change in their ability to change the income distribution. To be sure, some countries have increased their redistributive capacity -- France is a good example -- and others experience wild swings in redistributive capacity, probably related to big political conflicts -- note Bangladesh and Chile, the latter with a big bump around the time of Allende. But at best we can detect a long-term decline in redistributive capacity for the majority of cases (even if the decline is often slight); and often, after a decline, we see long periods of stability rather than change: countries settle into an "inequality regime," with some occasional big bumps which indicate new equilibria.

Note that in many cases the redistributive capacity of the state does not change even while inequality increases: thus, for example, while the net gini has increased over the past six decades in the USA and New Zealand, their capacity to affect the gini coefficient has remained approximately the same (New Zealand has been able to reduce the market gini by about 27%, though there's a slight downwards trend in this number; the USA by about 22%). The structure of their economies changed (by political action, in part), producing more inequality, but their redistributive capacity as states remained basically the same. To decrease inequality by redistribution in cases where the market gini increases substantially seems to require either a big political shock, or a long-run increase in state capacity.

There is another historical pattern that struck me as interesting: both levels of inequality and redistributive capacities seem to be highly correlated accross regions. Neighboring countries appear to have both similar levels of inequality and similar redistributive capacities. Linked economic and political histories seem to produce both the equilibrium level of inequality and the long-run redistributive capcity of the state.

Here, for example, we see the average redistributive capacity of states per region:

(Cases to the right of the solid line are states that on average made their income distribution more equal; the dashed line indicates the median redistributive capacity).

And here is a graph of net gini per region (all observations since 1960):

No great surprises here, perhaps: Sub-Saharan Africa and Latin America have been the world's most income-unequal regions over the last six decades, whereas Europe has been consistently equal - the home of both Northern social democracy and Eastern European communism, both of which have been able to keep the distribution of incomes relatively equal through explicit redistribution, though in somewhat different ways.

But perhaps this is of little importance. On one (vaguely Rawlsian) view, what matters is not the income distribution per se, but the ways in which it affects the prospects of the worst off in society. How much does it matter whether or not inequality declines in any given society, especially for the poorest? This will depend on the growth rate of the economy; high growth with declining inequality will be better for the poor than low growth with increasing inequality, though the outcome of the comparison is ambiguous for high growth with increasing inequality or low growth with decreasing inequality.

Now, it occurs to me that with the average income for these countries as well as their level of inequality, we can make an informed guess (technically, a wild guess) about the average income of various deciles for the years in which data is available. To do this properly would be too painful for a blog post, but I assume that empirical income distributions more or less fit a lognormal distribution (even though they fit more exotic distributions better, like the Singh-Maddala distribution or the generalized Beta distribution). With a little help from R, I can then simulate the average income of each decile of every country in the SWIID dataset. (Take a look at the code for the gory details. Also: this is a VERY rough and ready simulation, extremely inefficient to run and cooked up in a day. Do not take these numbers too seriously). We can then provide some vaguely informed answers to a Rawlsian question: which countries have most increased the prospects of the poorest over the last six decades?

The question admits of two more precise formulations, which we'll take on in turn: what countries have had the highest growth rate of income for the lowest deciles of the population? And second, in which countries do the poor have the highest incomes? (The first corresponds to a sort of dynamic version of the difference principle, which I find more interesting). Let's start with looking at the Rawlsian growth rate (the rate of income growth for the lowest decile, which we'll assume represents the group whose position must be maximized in Rawls' theory); the higher the long-run Rawlsian growth rate, the more the country fulfills the dynamic version of the difference principle. Though in theory the long-run growth rate of the economy as a whole and the long-run growth rate of the income of the lowest decile should perhaps converge, in practice they diverge, even over long time periods - some groups do well over some time frame, others do badly. Now, what we would actually want to know from a strict Rawlsian perspecive is the highest long-run growth rate of income for a representative person in the lowest decile of a given country relative to the potential growth rate of the whole economy (in other words, what degree of inequality would produce the highest income growth for that representative person, given the particular structure of that economy), but this is a counterfactual quantity we cannot estimate, so we'll make do with simulating the incomes of the lowest decile for the actual combinations of growth and inequality in existing economies. (I repeat my warning: this is only a simulation!)

First, we estimate the long-run Rawlsian growth rate (for countries with data going back far enough - so we drop countries that don't have long enough time trends, say at least 30 years):

The "Asian Tigers" unsurprisingly top the list: over the last six decades, Taiwan, Singapore, and South Korea had the highest (simulated) Rawlsian growth rate (in countries with at least 30 years of both GDP and gini data). South Korea and Taiwan are below average in inequality, which makes sense, but Singapore is not. Over the long run, in other words, a high enough growth rate of income seems to compensate for higher than average inequality. But one surprise among the top countries is Egypt - where the poorest decile, if we believe this simulation (and you shouldn't), had a pretty good run over many decades, despite Egypt not being considered a big performer in terms of its average per capita growth. At the bottom, by contrast, we find that Venezuela has essentially experienced zero Rawlsian growth over six decades (in fact, its long-run trend in regular per-capita annual growth is also zero). Though below average in inequality, its income has suffered so many ups and downs (mostly following oil price changes) that the trend is flat; no wonder Venezuelans eventually got tired of all their politicians before Chavez.

Now, there obviously is a correlation between Rawlsian growth and regular growth, as well as between Rawlsian growth and average inequality, but it is not perfect, simply because the Rawlsian growth rate is a function of both the average growth rate and the gini index by construction, and these two things are not perfectly correlated; a high enough growth rate in the whole economy can overcome a large gini coefficient to produce high Rawlsian growth rates and vice-versa. But it's worth noting that extremely high levels of inequality do appear to be associated with plain low growth over the long run, bad for both the poor and everyone else except perhaps tiny kleptocratic elites:

We can now repeat the exercise for the last 15 years and see how Venezuela stacks up since then:

As we can see, the Chavez years (up to 2010; the data does not tell us what happened for the last three years) were quite good for the poor, according to this simulation: the combination of declining inequality and relatively high growth rates (due in great part to rising oil prices) made Venezuela a top ten Rawlsian performer - better even than China, which also had torrid growth rates but increasing inequality during this period. To be sure, this good "Rawlsian" growth rate is only relevant if we ignore the equal liberties principle, which from a strict Rawlsian perspective  is meant to have priority over the difference principle; and increasing disregard of classic liberal rights during this period counts against Venezuela. (I vaguely wondered whether perhaps a "Rawls index" could be constructed, using data like the UDS to measure compliance with the first principle, fair equality of opportunity using the gini index, and the difference principle using the rate of growth of the income of the lowest decile; but since the two principles are supposed to be lexicographically ordered, a combined Rawlsian index would be pointless, useful only if we relax that assumption. Nevertheless, if we relax that assumption, then we would have to face the question of how much the improvement in the condition of the least well off ought to count against the decline in the "equal liberties" of the first principle; and I don't know of any good principled answer).

At the same time, it is interesting to note the countries at the very top are not precisely all left-wing governments; Azerbaijan, Mongolia, and Ukraine appear there. This may be because the simulations are risibly wrong (an important possibility), or the data are wrong; or simply that policies of the kind the Chavez government tried out are not the only possible ones to bring about growth in the income of the poor (and now, with high inflation, sporadic shortages, a large black market premium for dollars, and other problems, they don't look especially sustainable either). Nevertheless, the high Rawlsian growth rate makes it easy to understand why many of the Venezuelan poor felt that Chavez improved their position, regardless of how much responsibility we ought to attribute to his government for that outcome, or how sustainable its policies may be with lower oil prices.

Regardless, a good growth rate for the poorest decile matters: if inequality had remained at its maximum level during the Chavez years instead of declining but the growth rate had stayed the same, I estimate that the a representative of the poorest decile would have earned about $2000 less over the entire period than they actually did. We can call this quantity the "Rawls gap": the amount of income the poorest decile would have gained (or lost) in a given period had inequality remained the same as at the beginning of the period. Of course, since the growth rate would have been different had inequality remained the same, this is merely a fiction; we can't really estimate this counterfactual.

Nevertheless, just for fun, here is the Rawls gap for Latin America, per year:

This allows us to say that in Venezuela, the reduction of inequality that occurred during the Chavez period (assuming, per impossibile, that the growth rate would have stayed exactly the same had inequality remained at the 1999 level) gained a representative person in the poorest decile a total of about $1500-$2500 over 10 years, or about $200 per year, whereas the increase in inequality over the same period in Costa Rica cost a representative person in the poorest decile about $800-$1200 in income, or about $100 per year. This is nothing to sneeze at for the poorest decile (whose average yearly income is only about $3000 per year).

(It's kind of fun, though conceptually pointless and computationally expensive in my system given my crappy code, to calculate various Rawlsian gaps for arbitrary years and countries; for example, the "Rawls gap" for NZ is something like $2000 per year lost in income for the poorest decile if we assume the same growth trajectory but the level of inequality of the early 80s. Which of course we shouldn't - had inequality remained the same, the growth trajectory would have been different. As Adam Przeworski has said, everything is endogenous).

(We could also imagine even more exotic quantities, though I have no time to test them out here. Consider the Rawlsian compensatory growth rate, for example. This would be the growth rate that would compensate the poorest decile for an increase in inequality: if we want to say that some reform x would lead to higher income growth but higher inequality, then the compensatory Rawlsian growth rate is the growth rate where the income growth rate of the poorest decile at the higher level of inequality is identical to their income growth rate at a lower level of inequality but a lower overall growth rate for the economy; you would need a reform to produce at least the compensatory Rawlsian growth rate for it to be justified in terms of the difference principle. Which you may of course think is bogus).

Now, absolute incomes matter too; the difference principle in Rawls is not usually understood in terms of growth rates (though I think that should be the more natural interpretation). But the second version of the Rawlsian question above (where do the poor have the highest incomes?) has a much more obvious and boring answer: the Scandinavian countries, due to both generally high incomes and low levels of inequality due to high redistribution; and most of the countries at the top also score well in terms of the first principle (measured inexactly here by the UDS, which perhaps ought to be discounted a bit given recent developments in some countries). I include it here for completeness:

The roots of that ranking of countries are much older and deeper than this dataset allows us to see.

In theory, both the first ("equal liberties") and the second principle of Rawls' theory ("fair equality of opportunity" plus the "difference principle") ought to go together. In practice, however, Rawls himself thought that they did not always do so, though his reasons for thinking this were not always clear. Though I don't really have the tools to tackle the question of the relationship between liberal rights and the rest of the components of Rawls' theory properly (certainly not here), it looks as if we see a kind of inverted-U relationship in the data:

In other words, over the last half-century, the income of the poor has risen fastest under regimes that have not been on average highly democratic, but also has grown least in these regimes; non-democracy looks like a (potentially quite bad) gamble, though both democracy, long-run inequality, and the long-run growth rate of the income of the poor are probably determined by (or are a reflection of) some deeper social fact, like state capacity, which is not really susceptible to policy intervention. Whatever state capacity is (I have argued it is a kind of development of political technologies) it emerges out of political struggles that take a very long time to work themselves out with many tragic consequences along the way; and in any case the rate of improvement of state capacity is at times immeasurably low.

[Update 12/7/2013 - fixed a reference to a non-existing graph]

The Normativeness of Democracy

(Contains some work in progress, baroquely complex graphs to illustrate the obvious, rank speculation, and half-baked argument. It also continues this series on the quantitative history of political regimes).

Almost every country in the world publicly acknowledges the “normativeness” of democracy today. Democracy has become a sort of universally invoked standard, even though people vehemently disagree about its meaning. How do we know this? For one thing, almost every country in the world describes itself as a “democracy” in its constitutional documents. Using the data collected by the Comparative Constitutions Project, we can see that as of 2006, only 20 of 184 countries with some kind of written constitutional document did not describe themselves as democratic:

Fig. 1 Countries that do NOT describe themselves as democratic in their constitutional documents

This understates the universality of the norm. As we can see above, many of the countries that do not explicitly mention the word “democracy” in their constitutional documents are countries whose public culture nevertheless asserts and assumes that they are long-standing democracies - a judgment typically confirmed by democracy indexes like the Unified Democracy Scores. (There is in fact a slight negative correlation between the current or the long-run degree of democracy, as measured by such indexes, and whether or not the country calls itself a democracy; almost every country in the map above is represented by a blue dot, which indicates that observers generally regard them as democratic).

At any rate many of the countries that do not describe themselves as democracies in their constitutional documents have constitutions dating back to a time where the word “democracy” didn't carry the positive associations it does today, like the USA and the Netherlands; and some nevertheless use words that are effectively equivalent to the word “democracy” today, even if in the past their usage differed, like the word “republic”. Thus the USA constitution guarantees a “republican” form of government to every state; Singapore and Yemen explicitly describe themselves as “republics” in their constitutional documents, and Yemen asserts its adherence to a principle of “political and partisan pluralism”; the Japanese constitutionstates that “sovereign power resides with the people” and that “government is a sacred trust of the people, the authority for which is derived from the people, the powers of which are exercised by the representatives of the people, and the benefits of which are enjoyed by the people;” Jordan stresses that its monarchy is “parliamentary;” the Netherlands, Belgium, Monaco, Denmark, and Norway all describe themselves as “constitutional” monarchies; and of course almost all of these constitutions guarantee some form of universal suffrage. Indeed, of the countries listed above, only the absolute monarchies - Oman, Saudi Arabia, and Brunei - really refuse to pay any lip service to the norm.

Moreover, the assertion of “democracy” in constitutional documents is almost always accompanied by the assertion of the classical “liberal” norms: freedoms of speech, expression, religion, association, press, and the basic equality of all people. The constitutions of the most repressive countries all proclaim such freedoms. Let's take the basic freedoms of association, speech, religion, and assembly, as well as the norm of equality before the law. Almost every constitutional document in the world (over 90%!) asserts all five of these; and among those countries that don't, most proclaim their allegiance to at least four of these. Only two countries (New Zealand and Libya!) failed to mention any of them as of 2006:

Fig. 2 Constitutions missing one or more "liberal" freedoms as of 2006

In the case of New Zealand, this is basically an artifact of the choice of “constitutional document” coded by the Comparative Constitutions Project - the New Zealand Constitution Act, last revised in 1986, is a purely structural document, setting out the powers of parliament and the governor general and describing various other institutions, while other documents, including the Treaty of Waitangi, play a more important role in setting out the important normative commitments of the country. (Many people would anyway say that New Zealand does not have a written constitution; the Constitution Act is simply one part of the constitutional law of the country, and possibly not the most important part). In the case of Libya, the post-Qaddafi interim Libyan constitutional declaration, article 14, explicitly asserts all of the freedoms that the Qaddafi 1977 amendment to the 1969 “provisional” constitution failed to mention. (Interestingly, the 1969 Libyan provisional constitution did mention some of these rights, but they were apparently excised from the 1977 revision).

I find this striking. The old saw about how “even the constitution of the Soviet Union” proclaimed freedom of the press, expression, association, and the like was not only true of the Soviet Union; it is true today of practically every country, however repressive; indeed, many of the countries that fail to expressly list such rights and freedoms in a constitutional document nevertheless affirm them in their public culture. (Note that all countries not listed in the graph above explicitly assert all five liberal freedoms in their constitutional documents, whenever such documents can be identified, which is almost everywhere today; North Korea, for example, affirms all of them). Yet it is obvious that the mere affirmation of these principles does not imply that they are honored in any way shape or form, and in some places the assertion can only be taken as mockery. If we take the UDS score as a very rough measure of how likely it is that these rights are honored in practice, with higher (“more democratic”) scores indicating more likely respect for these liberal freedoms, then we are forced to conclude that there is basically no correlation between expressing normative support for such freedoms in constitutional documents and actually protecting them.

This might seem unsurprising; a cynic might say that I've only rediscovered the obvious impotence of constitutional restraints in the absence of supportive social and political realities. But it is nevertheless interesting, to my mind, that there is such a widespread need to assert these particular normative commitments, even as they are routinely violated, or interpreted in such radically restrictive ways as to render them politically meaningless. Among authoritarian elites, only the House of Saud and the Sultan of Brunei appear to have the courage of their convictions; everyone else hides behind a banner of rights and liberties.

Nevertheless, only some rights within the larger universe of potential normative claims are universally asserted; if we take a look at the full list of rights that the CCP codes - ranging from freedoms of expression, opinion, and association to the right to bear arms or be granted asylum to socioeconomic rights like the right to own a business, to strike, to healthcare, or to a specific standard of living (about 64 in total) we find a small core of liberal rights that basically everybody asserts (plus a right to own property, in uneasy balance with a right of the state to expropriate it, typically with compensation) and a larger penumbra of other rights, different sets of which are asserted by various sets of countries. In the graph below, each word represents a particular rights provision tracked by the CCP project, surrounded by colored dots representing the countries whose constitutions contain that provision. (A list of all of these provision is available in the codebook here). The number near the word represents the proportion of countries that assert the provision (for example, 90% of all countries assert a commitment to protect freedom of association, “assoc”, in their constitution); the color of each dot shows the UDS score of the country as of 2006, where blue indicates “more democratic” and white indicates the dividing line, more or less, between democracies and non-democracies.

Fig. 3: Rights and countries, fireworks mode

(You come here for the intellectual fireworks, right? There, some fireworks). A perhaps more rational (but less fun!) way of visualizing the same data is this:

Fig. 4: Proportion of constitutions affirming particular rights

(Click to enlarge. Red lines indicate where 50% and 80% of the constitutions of the world explicitly affirm a particular provision; the color of the dot represents the average UDS score of countries that endorse a particular right. It was interesting to note that the right to bear arms appears to be unequivocally endorsed by only about 1.5% of the world's constitutions - the USA, Mexico, and Guatemala).

We might thus say that the “liberal” rights and the associated idea of democracy appear to have a good claim to represent a sort of global “overlapping consensus” in Rawls' sense - rights that are publicly accepted for diverse reasons in very different societies- and may serve as a basis for normative judgment everywhere. (Incidentally, this the case not only for public constitutional documents, which may be thought to be elite-imposed and not always faithful reflections of the normative aspirations of the broader society; though this requires another post, public opinion in most countries also seems to overwhelmingly support democracy and many “liberal” ideals [ungated], at least in the abstract, even if it is not always clear what this actually means in practice. Talk, of course, is often cheap, and abstract support for democracy and liberal freedoms does not necessarily translate into genuine concern for them.) Other rights, however, are still objects of normative struggle at a global level; they are not universally accepted.

But though there is much less consensus about these other rights, it is nevertheless striking that public affirmation of any set of these rights is not obviously clustered in particular societies either. It isn't always clear why a society chooses to “constitutionalize” a particular right, and publicly affirm it in its highest legal document; but whatever the case, democracies and non-democracies are about equally likely to endorse a given right in their constitutional documents. As we see above, the average UDS score of countries endorsing any particular right is pretty much the average level of democracy in the sample, at least for provisions endorsed by, say, more than 10% of the world's constitutions.(We can also see this by noticing that of the fireworks above are especially red or especially blue, save for rights explicitly endorsed by very few countries, like the right to bear arms - USA, Mexico, Guatemala - or the provisions specifying that law contrary to religion is void). Furthermore, there is no correlation between the number of rights provisions endorsed by a constitutional document and either the contemporary or long-run level of democracy, as measured by the UDS score or the cumulative UDS score. In fact, constitutions in general seem to be fairly similar to one another; and to the extent that particular sets of constitutions cluster together (grouping together countries that affirm similar sets of rights) these clusters do not correspond to obvious cultural, political, or other groupings.

One way to see this is as follows. (Please indulge my taste for complicated graphs). We take the list of rights and duties coded by the CCP and calculate the matrix of “distances” between them - essentially, we calculate how similar each constitution is to each other along that set of dimensions, using the Gower similarity coefficient, where 1 means the two constitutions are exactly alike (they affirm the exact same rights) and 0 means they are completely dissimilar. We can then use this distance matrix to plot the world's constitutions as a network and visualize their clustering patterns; highly similar constitutions should cluster together (they are less “distant” from one another). And indeed, we can see some patterns (community discovery algorithms suggest the graph below has about 4 big components when we include all links), but these patterns do not correspond to any obvious groups, like democracies or dictatorships, or poor and rich countries. Indeed, the groups obtained in this way can seem downright perverse, placing, say, Germany and Egypt closer together than Germany and the USA:

Fig. 5: Network of similarities among constitutions (rights provisions only, 85% similarity and up)

Or, alternatively, take a random constitution from a democracy (as measured by an UDS score in the top 33% in 2006) and a random constitution from a dictatorship (as measured by an UDS score in the bottom 33% in 2006) and they will share, on average, about 60% of all rights provisions tracked by the CCP project (and about 80% of the basic liberal democratic freedoms of assembly, association, etc.); take two random dictatorships or two random democracies and they will share similarly about 60% of their rights provisions (and 80% of the basic liberal freedoms). The same is true if you look at the “duties” provisions of constitutions - e.g., whether the state has a duty to provide work, or citizens a duty to work or serve in the military. Or, indeed, any other set of provisions tracked by the CCP; it seems difficult to find any dimension - descriptions of executive power, electoral provisions, etc. - along which the constitutions of more or less democratic societies, or societies in different regions, or at different levels of development, appear to be systematically different (any two random constitutions are about 65% similar, taking all dimensions together). In other words, the normative self-presentation of societies whose power structures are widely different (at least as measured by standard indexes) is pretty much identical; if I'm right, you could not systematically say much about the kind of power structures in a society by looking at its constitution.

(At this point, this thought goes through my head: “Are my methods unsound? I see no method at all, Mr Marquez”)

What might explain this “normative convergence”? The point, it is worth emphasizing once again, is NOT about the effective enforcement of constitutional norms; I take it for granted that such norms -specifically, the norms granting individual rights to citizens, of whatever kinds- are only spottily effective in most places, even in many “democratic” countries, though I think it is reasonable to assume that countries conventionally held to be democracies (as measured by the UDS) will tend to enforce whatever rights appear in their constitutions slightly more effectively than the average non-democracy (if perhaps not much more effectively, and with many exceptions). I'm interested, instead, in the “normative mimicry” on display here, and the process through which some norms achieve near-“fixation” in the population, despite what we might call their fictional status in many cases.

Now, before you accuse me of being willfully obtuse, I am aware of the obvious explanation: modern societies, the story goes, required a new “basis for legitimacy” after the breakdown of traditional forms of legitimacy. Norms of popular sovereignty and individual rights come to replace earlier “legitimizing” norms; and so all regimes now “legitimize” their power by appealing to these norms. But I'm not sure that this doesn't simply restate the problem. Why these particular norms and not others? And why would appeal to “fictional” norms - norms that are known not to have any substance on the ground, so to speak - legitimize anything (in the sense of increasing the baseline level of support for a structure of domination)? It's not that there are no answers to these questions; it's just that the appeal to legitimacy is question-begging if what we are trying to explain is how the norms became dominant in the first place, even when they have minimal impact on what happens in day to day life.

There are more and less “optimistic” stories one could tell about this process. An “optimistic” story could say that there was a sense in which the norms of liberal democracy and its associated freedoms became increasingly appealing to people throughout the world over the last two centuries, while alternative norms became less so. (One might here appeal to increasing literacy, capitalist development, the breakdown of local solidarities, etc. to explain the formation of modern subjectivities; whole libraries have been written on these topics). Normative change outstripped social change; and every political regime now feels compelled to pay at least lip service to these “new” norms, if only because not mentioning them exerts some negative pressure on their survival prospects, perhaps by making those subject to it needlessly dissatisfied. By the same token, this story might continue, the mere existence of the norm puts pressure on governments to live up to their highest commitments, and enables dissatisfied people to coordinate their claims; thus Chinese activists, for example, have (on and off) appealed to the party to enforce China's own constitutional norms guaranteeing basic freedoms of speech, association, etc., and perhaps eventually they will get somewhere. Accordingly, even if normative change feels insignificant at first, it can be utterly momentous in the long run - like a force that exerts only a slight pressure, but does so continuously over the very long run and so ends up accelerating a great mass to huge velocities.

While this story is probably not entirely incorrect, it seems to me that the problem here is that for a norm to have any kind of ability to raise the baseline level of support for a political structure, it needs to be not only widely recognized as a normative standard, but credibly asserted by those in power; and many of these norms are not. (It seems absurd to me to think that the mere assertion of freedom of speech in the North Korean constitution can possibly fool anyone who doesn't already want to be fooled for other reasons, to take only an obvious example). Moreover, it seems clear that many of these norms are liable to lose their force as they become globally dominant simply as a result of adaptation on the part of groups adversely affected by them. There was a time, for example, when it was a matter of live controversy who should be admitted to the franchise, whereas nowadays most adults everywhere are enfranchised, even if their votes are utterly meaningless, since powerful groups have adapted to the mere existence of elections (if not necessarily to the possibility of actually fair elections). Similarly, it may be that as legal freedom of speech becomes increasingly unlikely to genuinely threaten powerful interests, the more easily it comes to be accepted as a global norm. Successful adaptation by groups that are disadvantaged under particular norms reduces their propensity to produce conflict; and the global dominance of a norm can thus mean either that it is ripe for struggles to give it substance (let's turn the fake democracy into a real democracy) or that it has been hollowed out, and live conflicts have relocated to other normative arenas (the right to healthcare, or to a standard of living, or to bear arms, or to enforce one's religion on others, etc., rather than suffrage, etc.).

There are also other complications. Suppose that particular norms become entangled with markers of status; to be a “proper” country, with a “proper” constitution involves asserting some of the norms that powerful countries profess to affirm. As long as the norm is merely one of the marks of status tied to a specific collective identity, it can be asserted by most people in an entirely fictive way. The norm then appears as a sort of ornament, one aspect of a collective identity expressing “far” values, while being ignored or rationalized away in concrete situations. (The modern USA is “the land of the free” to most of its citizens irrespective of particular facts about freedom in the USA; and the idea that Venezuelans have “the best constitution in the world” is entirely unaffected, for most Venezuelans, by the fact that it is routinely ignored). On this view, it is precisely higher-status countries that have the most freedom to mention or not to mention particular norms, which is more or less in accord with some of the data above, though I have not checked properly; and “new” norms should come from relatively peripheral countries with leaders intent on raising their status (e.g., Venezuela, whose constitution is chock-full of rights and institutional innovations, however unenforced). (Incidentally, we know that in fact many “democratic innovations” first emerged and were developed in peripheral countries, not major powers). The power of the norm comes here less from the content of the norm - as in the optimistic story - than from its association with other markers of status. I suspect similar things happened in the more distant past; as James C. Scott notes in The Art of Not Being Governed, the symbols of absolute monarchy were often adopted by peoples who had hardly “states” at all: every two-bit chieftain claimed to be a universal emperor.

The global dominance of “democratic” norms in this "fictional" sense complicates our efforts to make sense of events like the Arab revolutions. Were these revolts “for” democracy? People sometimes argue that the revolts were not “for” democracy insofar as many protesters didn't make “democracy” their principal demand; instead, people wanted jobs, respect, dignity, and many other things. But we need to take into account the fact that the Arab republics explicitly endorsed democracy - the Qaddafi constitution made a huge deal of its participatory democratic character, for example - yet the norm was without substance. The revolts have sometimes attempted to give substance to the norm, but sometimes they have chosen different, more contested normative terrains - over the role of religion in public life, for example - where a norm is not yet universally accepted. This does not mean that “democracy” was not valued; it may mean merely that it was not always understood as something that required normative defence, or as a terrain where fighting over meaning was likely to lead to any interesting places, since everyone already agreed on democracy as the standard, though they disagreed in how exactly to give substance to it. Anyway, more of this would become clearer if we had a better sense of how it came about that these “liberal” rights became so dominant as normative fictions - when and where they first became publicly affirmed throughout the world. But I've run out of steam, and this post is already long enough. More later on the more vexed question of “culture” and democracy, perhaps…

Code for replicating the graphs in this post (plus some additional stuff) is available in this Gist (five files, including one with auxiliary functions and some geocoded country codes). You will also need access to the public CCP data.

The Good, the Bad, and the Ugly: More on Benevolent, Malevolent, and Unconstrained Regimes

(Quick graphical follow up to this post on malevolent democracies and benevolent autocracies; part of this series, asymptotically approaching 2. For discussion of the measures of physical integrity, regime types, democracy, and executive constraint used here see the previous post. Because I just can't help myself).

After finishing the previous post, a clearer way of presenting the idea of a "benevolent" regime occurred to me. Essentially, we can classify all regimes along two dimensions: the degree to which the executive is constrained by formal institutions, and the degree to which the state (directed by the executive) engages in killing, political imprisonment, torture, and so on. Using the CIRI data on the protection of physical integrity rights, the Polity IV measure of executive constraint, and the DD measure of political regimes (by Cheibub, Gandhi and Vreeland) we then get the following four-fold classification:

Benevolent regimes are on the upper left hand quadrant: here, rulers are formally unconstrained but nevertheless have, on average over the last 30 years, respected the physical integrity rights of their subjects. These tend to be relatively wealthy, as we can see, and many of them are absolute monarchies: the Qatar of al-Thani (2001-2008), the Oman of Sultan Qabus ibn Sa'id (1981-2008), Swaziland under various monarchs, Bhutan under Sigme Wangchuk (I've labeled the top 5% of the regimes by the benevolence measure, though they may be hard to see, and some are missing from the plot because they don't have GDP data). But many of them are poor countries (even if their GDP figures are occasionally inflated by oil discoveries, for example) and frankly a bit surprising: Gabon under Bongo (1981-2008), Burundi under Buyoya (1981-1991), Malawi under Banda (1981-1993), for example. All of the latter were rulers who consolidated their power long before 1981, I think; so perhaps with a longer dataset we would get somewhat different results. But still, it is worth noting that "the good" are sometimes long-established autocrats.

The bad are typically regimes that hold elections and have formally constrained executives, but where public opinion and the political class is indifferent or even supports violating the rights of various groups of people: Colombia, India, Israel, Indonesia, all make appearances here, as well as South Africa under Apartheid and Peru during the Sendero Luminoso years. The constraints have stopped working with respect to particular groups - poor peasants in Colombia, native peoples in rural areas in Peru, people who are thought to be associated with separatists in the Philippines, Palestinians in Israel. 

The ugly are the usual suspects: your typical unconstrained, megalomaniacal dictator, like the regimes of Gaddhafi in Libya, Kim Jong-il in North Korea, Milosevic in Serbia, Mobutu in Zaire, Marcos in the Philippines, Galtieri in Argentina, al-Bashir in Sudan, Hoxha in Albania, Taylor in Liberia, and so on. It's the dismal roster. Interestingly, Buyoya of Burundi appears again here (1996-2000) among the unconstrained  - a striking illustration of the fragility of mere benevolence. 

The constrained are the boring regimes - not without problems, to be sure, but about as well-functioning as we usually get. (Though some, of course, may note that constrained regimes "internally" does not mean that the regime will be constrained "externally"; as I mentioned in the previous post, threats to the state seem to turn constrained regimes bad). 

Here, just for the hell of it, is a list of countries ranked by their average degree of benevolence; the boxes tell you were how much their "benevolence" has varied over time, and the dots are the outliers - years where their benevolence has strayed far from the mean:

Fig. 2. Regime rankings by benevolence. Dashed lines identify the USA, Venezuela, and New Zealand

We can see at a glance that "benevolence" is mostly a phenomenon of autocracy, though a few democracies have above-average levels of it; and that malevolence is mostly a democratic phenomenon. Yet benevolence among autocracies varies a bit more than among democracies; your average benevolent despot is not extremely reliable, perhaps. (Note also how Burundi is represented at the top and the bottom of the scale, with two different regimes). It is mostly countries in the middle of the distribution that have consistent records of protecting rights, though they are not unblemished.

Here a final picture ranking countries by their average levels of protection of physical integrity rights (thickness of the line represents democracy level; color represents level of protection - darker blue is better; each colored line shows changes from 1981 to 2008): 

Fig. 2: Ranking of countries by avg. levels of physical integrity protection, showing changes during the 1981-2008 period. Dashed lines identify the USA, New Zealand, and Venezuela

New Zealand has done very well during this period (by this measure at least), but the USA ranks 35 out of 167, with a big dip at the end of the period, and a spotty record (note the light blues in earlier years); and Venezuela's state has been going bad since 1989, with the Caracazo. More dictatorships appear at the bottom of the scale, but also many democracies. Interestingly, the end of the cold war seems to have produced improvements in the protection of rights in very few countries; at a glance, only Nicaragua, El Salvador, and Albania stand out. But New Zealand is only in the middle of the benevolence ranking; it does well because its constraints have worked, not because its rulers are unusually benevolent, though perhaps its constraints have worked because public opinion has been reasonably enlightened, and public opinion has been reasonably enlightened because it has faced no big conflicts in the recent past.

(Some messy code that produces these and other graphs is here).

Of Malevolent Democracies and Benevolent Autocracies: A Very Short Quantitative History of Political Regimes, Part 1.9325

(Continues my occasional series on the history of political regimes, part 1.9325. Lots of charts and graphs and a slideshow, using the Cingranelli-Richards (CIRI) human rights dataset and the Political Terror Scale.)

About a month ago, Reed Wood over at the blog Political Violence @ a Glance expressed doubts that there had been much if any meaningful improvement in the extent to which states engaged in torture, beatings, etc. over the past three decades. Neither the Political Terror Scale (measuring the degree to which states engage in torture, political imprisonment, or political murder) nor the Physical Integrity Index of the CIRI dataset (which measures more or less the same thing in a somewhat different way) show any improvement over the last three decades or so, despite the fact that (as we have seen here, here, and here in much greater detail) the world is a more "democratic" place today than three decades ago:

Figure 1. Global mean of the CIRI Physical Integrity Index,  1981-2010 (higher means more protection of physical integrity rights, on a 0 to 8 scale) 

Figure 2. Global mean of the Political Terror Scale,  1976-2011 (higher means more state use of torture, murder, and other physical integrity violations, on a 1 to 5 scale) 

Figure 3. Global mean of the polity2 score,  1976-2011 (higher means more democratic, on a -10 to 10 scale) 

If anything, a slight worsening trend in the extent to which states engage in torture, killing, and so on is detectable here, despite the increase in democracy over the same period. This piqued my curiosity; what is going on here? And what has been the relationship between political regimes and the protection of basic "physical integrity" rights, historically speaking?

Neither the CIRI dataset nor the Political Terror Scale are ideal for answering these questions in sufficient depth; for one thing, their data collection starts just after "peak authoritarianism" in the mid 1970s, and hence misses much of the consolidation of authoritarian regimes in the 50s and 60s. Moreover, it is very likely that the lack of improvement we see above is at least partly an artifact of better reporting and a broader understanding of what counts as a violation of physical integrity by the state, as Anne Marie Clark and Kathryn Sikkink argue in a forthcoming piece. But they can still help us understand broad correlations between political regimes and the malevolence (or restraint) of the state.

Using the CIRI data and a typical measure of democracy (the Polity IV scale, discussed here in more detail), the first thing we note is that, though the overall trend in the protection of physical integrity rights is negative across all regime types, democracies have had on average a better record than other political regimes over the last 30 years:

Figure 4: Trend lines for physical integrity rights by regime type. "Democracy" is defined as a polity score greater than 6; "Autocracy" is defined a polity score lower than - 6. Each point represents a country-year; points are jittered to avoid some overplotting. 

Though we see dictatorships that appear not to engage in torture, killing, and so on, in every year, as well as democracies that do engage in such practices, on average democracies score about 2 points higher in the CIRI Physical Integrity Rights index than autocracies and "anocracies" (the Polity IV term for various hybrid regimes; the picture does not change if we use the Political Terror Scale instead). Since the CIRI index is additive over four measures of state malevolence - extrajudicial killings, disappearances, political imprisonment, and torture, each of which is scored as 0 if the violations are judged to be frequent and widespread, and 2 if they are judged not to have occurred during a particular year (see here for more details) - we could say that democracies on average have committed one fewer crime than autocracies over this period. (Democracies: statistically less criminal than autocracies for over 30 years!). Indeed, nearly 80% of all regimes that receive a perfect 8 in the CIRI index are democracies, while nearly 75% of the regimes that receive a 0 in the index are autocracies or anocracies:

Figure 5: Distribution of regimes accross CIRI index categories. 8 means a high level of protection of physical integrity rights; 0 means widespread violations. "Interruption" includes Polity categories for breakdown of central authority, foreign occupation, and transitional forms.

Moreover, it is also worth noting that the average difference between democracies and autocracies has not changed at all over the entire 30 year period, even as both democracies and autocracies seem to be becoming worse (i.e., more likely to engage in torture, killing, etc.). To me, that looks like prima facie evidence that the lack of improvement in these indexes is at least partly a reporting artifact, though other stories are possible. For example, suppose that the least malevolent autocracies are most at risk of turning into democracies. But they do not immediately become "high quality" democracies; political competition restrains states slightly better than before, but still not well. As more countries become democratic, the average malevolence of both democracies and autocracies should increase - in the first case due to the influx of low-quality democracies into the population (dragging the mean down) - and in the second case due to the exit of less-repressive autocracies from the population. I don't know that this story is correct, but it's worth considering, and some of the trends I discuss below are consistent with this relationship. In particular, if the story is correct, we should see a (slight) strengthening of the correlation between measures of democracy and measures of physical integrity over time - and in fact we do see this.

At any rate, the correlation between democracy and a lower average level of state malevolence remains striking, if perhaps unsurprising given common ideas about democracy. But couldn't it be the case that the correlation is built into the measure of democracy we are using here? Though the Polity IV measures are basically institutional (conceptualizing democracy as a variety of political competition, without assuming much of anything about whether or not democracies are more or less malevolent state forms), it may still be the case that they assume too much. Perhaps coders tend to give "nice" regimes higher scores; some of the political competition categories in the Polity IV index are notoriously opaque. Yet the correlations are the basically the same if we use the most minimal measure of democracy we can think of (the dichotomous DD measure, discussed here, which defines democracy as a regime where the leadership of the state is selected through competitive elections and nothing else):

Fig. 6: Distribution of democracies and dictatorships across CIRI scores. Dichotmous measure of democracy from the DD dataset by Cheibub, Gandhi, and Vreeland.

As before, it looks like most "benevolent" states (those scoring high in the CIRI index of physical integrity protection) are democracies, and most "malevolent" states are dictatorships (or more precisely, regimes where the political leadership is not selected via competitive elections; whether we want to call these regimes "dictatorships" is basically a question of nomenclature). (Results are basically identical if we use the PTS). But there remain a good number of elective regimes who score very low on the index, as well as a good number of non-elective regimes that score high on this index. Let's call the former malevolent democracies and the latter benevolent autocracies. What can we say about them?

Consider first the distribution of CIRI scores across regime types (using the DD measures of regime types):

Parliamentary democracies (59 of them in the dataset, nearly 1000 country-years in total with data on physical integrity protection) and mixed presidential/parliamentary democracies (36 of them, about 500 country-years) are the clear winners here - the least overtly criminal regimes over the past 30 years. Monarchies, however, (14 of them in the dataset, for about 300 country-years) did quite well; few of them appear to have engaged in any significant malevolence during the 1981-2008 period (and most of it was concentrated in Nepal, Morocco, and Saudia Arabia during this time). Indeed, the mean level of "physical integrity rights protection" in monarchies is nearly as high as in parliamentary democracies, and higher than in presidential democracies, which have been the worst of the democratic regimes. Civilian dictatorships appear just as likely to be "good" as to be "bad," and military dictatorships are the worst of the bunch. No non-democracy (monarchic or otherwise), scores a perfect 8 average for the period, whereas some democracies do (mostly small places like Iceland or Tuvalu, or very new democracies that have not yet have time to besmirch their records), though of course a number of democracies score very low too. (The rankings of regime types don't look any different if we use the PTS instead of the CIRI index).

Perhaps more interesting is to look at the most benevolent autocracies and most malevolent democracies over the period. The color of each glyph in the map below represents the average level of the CIRI index for the years for which data exists; the size of the glyph represents the number of years with data (ranging from 1 to a maximum of 30); and the shape represents the average Polity2 score over the 1981-2010 period, split into three broad categories: circles represent countries that have been mostly democratic over the last three decades; triangles represent countries that have been mostly "anocracies" (hybrid regimes); and squares represent countries that have been mostly autocratic. So if the correlation between regime type and the protection of physical integrity rights were perfect, we would expect squares in the map below to be red, circles to be blue, and triangles to be light colored. Red circles thus represent malevolent democracies, and blue squares benevolent autocracies (I've labeled the most malevolent democracies and the most benevolent non-democracies below):

Fig. 8. Average levels of physical integrity protection, 1981-2010, by regime type as defined by the Polity IV score

The top benevolent non-democracies by this measure (average CIRI index greater than 6, average polity2 lower than -6) are a mixed bunch: Suriname, Poland, Hungary, Croatia, Gambia, Benin, Gabon, Qatar, United Arab Emirates, Oman, Taiwan, Singapore, and Fiji. Among these, Oman[!] has the best overall record, with an average CIRI score of 7.17, better than the USA average for the period. Some of these regimes are basically electoral but noncompetitive regimes, such as Singapore; others democratized substantially during the period in question, though reports of torture or political imprisonment appear never to have been common even during more authoritarian times, or were concentrated at the beginning of the period (Taiwan, Poland, Hungary); and others are rich gulf monarchies (Oman, UAE, Qatar). Perhaps the most surprising cases are Gambia, Benin, and Gabon, none of which appear to have engaged in much direct political violence against their own citizens (at least none that was noticed by the State Department or Amnesty International, the ultimate sources for the CIRI index), despite being very poor countries. Happy autocracies, pace Tolstoy, are not all alike.

The top malevolent democracies by this measure show more commonalities: El Salvador, Colombia, Venezuela, Brazil, South Africa, Turkey, Israel, and India. These are almost all countries that faced or face substantial internal conflicts - the FARC insurgency in Colombia, Kashmir and Nagaland in India, the conflict with the Palestinians in Israel, conflicts with Kurds and between the secularists in the military and more religious civilian forces in Turkey,very sharp class conflicts in Venezuela, Brazil, and El Salvador. "Internal" threats to the state turn democracies bad. Just look at the CIRI graph for the USA for the period, and check out the dip after 2001 for further evidence:

Fig. 9. CIRI index of physical integrity rights for the USA, 1981-2010

And all this time the USA received a polity2 score of 10 - the highest possible.

Obviously, these overall judgments have to be taken with a grain of salt; as the cases of Poland and Hungary show, the cutoff for the dataset may mean that more repressive periods are excluded from consideration, or it may mean that they are included when there has in fact been a qualitative break in the nature of the state. Furthermore, there may be other factors that are related to the level of physical integrity protection; the level of economic development, inequality, and the rate of economic growth all come to mind, among many other possibilities (some of which are explored below).

One may think that the key factor ensuring that the regime is not criminal is the degree of executive constraint, not the whole degree of democracy. So let's define a benevolent regime as a regime where the executive is highly unconstrained but nevertheless does not act in a criminal way (though it could do so with impunity) and a malevolent regime as one where the executive is so constrained but nevertheless is not prevented from killing, torturing, or imprisoning for political beliefs at least some of its citizens, perhaps because the people who constrain the executive are complicit in its criminality. Now, Polity actually includes a measure of executive constraint, which is highly correlated, but not perfectly, with the CIRI index. (Better, in fact, than the overall democracy measure). We can then ask: are there truly constrained malevolent democracies? Or truly unconstrained benevolent autocracies? Using the DD measure of regime type and the polity measure of executive constraint we can produce an independent "index of benevolence" (essentially, we multiply both, after reversing the exconst measure, and take the square root) - higher is more benevolent, with a median of 4:

Average degree of benevolence/malevolence, by country and regime type (as measured in the DD dataset)

As we can see, democracies are typically constrained but not benevolent; almost all of them score lower than the median of benevolence. The top 10 "benevolent" regimes are all non-democracies (except for Bhutan for a few years, which is a hard case), and we have to go down to Mali to find a regime coded as democratic by DD that also had a relatively unconstrained executive and a reasonable record of not torturing, killing, or imprisoning its citizens. By the same token, most dictatorships overperform a bit; their records are better than the degree of executive constraint would lead one to expect.

The top 10 benevolent autocracies by this method are similar to the ones identified above. Many are absolute monarchies - Qatar, Oman, Swaziland; constrained, perhaps, more by tradition and culture than by formal institutions, as Victor Menaldo has argued (ungated). Some are expected, though still puzzling (why so restrained?): Singapore, Hungary during the last decade of the communist system. Others are very unexpected - many very poor African autocracies - Gabon, Benin. At the other end of the scale we find that malevolence (executive constraints plus torture and imprisonment) is almost always a democratic phenomenon; India, Israel, Colombia, and Jamaica bring in the bottom of the table. When unconstrained dictators do these things it's expected, but when democracies do it it's malevolent.

A temporal view may be interesting too. In the slideshow below, the size of each dot represents the actual level of protection of human rights, the color of each dot represents the naive deviation from the expected level of protection of human rights given the polity score, and the shape of the dots represents the regime type. So red dots are more malevolent than expected given the polity score of the country and blue dots more malevolent, while white dots are at the expected level of human rights protection. The measure for the deviation here is not empirically derived - it's not the residual of a regression of the physical integrity index on the polity score - but normative; a democracy with a perfect polity score that does not engage in torture, killing, and so on is not "benevolent" but merely doing its job properly, whereas a completely unconstrained dictatorship with a polity score of -10 that does not engage in torture, killing, and the like of political opponents is being "merely" benevolent. Hence dots representing democracies with a score of 10 never look blue on the map, and dots representing dictatorships with a score of -10 never look red. I've labeled the countries that have the largest deviations from a simple naive relationship between polity2 and the level of protection of physical integrity.

(Best viewed in full screen by clicking on the link on the lower left corner). Note how the period starts with a lot of benevolent autocracies (including a large number of African countries whose polity score belies the relative benevolence of their states by this measure) and democracies that generally respected human rights. As the nineties come along, there are many new democracies that "underperform"; from a blue and white world (with an even split between benevolent autocracies and democracies that do their job), we come to a world that is mostly pink (with many new and underperforming democracies), consistently with the selection hypothesis mentioned above. Things then take a sharp dive in the aftermath of September 11; most democracies - including most established ones in North America and Western Europe - become more malevolent in the years after 2011. In some cases we can easily point to the specific events that turn countries red in the map: the Sendero Luminoso years in Peru in the 1980s, the Caracazo in Venezuela in 1989, the endless conflict with the Palestinians in Israel. But though many countries in the map appear as benevolent autocracies or as malevolent democracies for short periods, most countries seem to settle to their expected level; both benevolent autocracy and malevolent democracy seem to be fragile, though benevolent autocracy is more common than malevolent democracy.

It is also worth looking at how the relationship between democracy and human rights protection varied over this period. We simply fit a simple linear model regressing the Physical Integrity Index against the Polity2 score for each year, and look at how the coefficient for the polity2 score has changed over time:

Fig. 11: Coefficient of polity2 in the model PHYSINT ~ a*polity2 + b estimated for each year. Shaded areas represent the 95% confindence interval for a.

The picture is suggestive of a structural break in the relationship between democracy and state malevolence with the end of the Cold War. States that had apparently been more autocratic than their record of benevolence suggested suddenly found their expected level of democracy, consistent with the hypothesis mentioned above. But not every country has benefited from increases in democracy. Among countries where we observe changes in their level of democracy in this period as measured by the polity2 score only about half of them (56% or so) seem to have experienced changes in the level of state malevolence that are in the right direction. In other words, it is only in about half the cases in the sample that increases in democracy are (statistically) associated with greater state benevolence (and vice-versa: decreases in democracy are statistically associated with greater state malevolence); in the other half, increases in democracy are associated with greater state malevolence (and vice-versa), though the magnitude of the association appears to be small in most cases:

Fig. 12: Magnitude of the relationship between polity2 and CIRI, per country, 1981-2010. Lines show 95% confidence intervals for the polity2 coefficient

In countries to the left of the red line, increases in democracy were associated in this period with more state malevolence (or vice-versa, i.e., increases in autocracy with more benevolence); in countries to the right, increases in democracy are associated are associated with more state benevolence (as we would naively expect). The striking thing here is how little of a pattern there is; though there are some slight regional associations (democratization appears to have been more correlated with state benevolence and vice-versa in the Americas than elsewhere), and some events are not captured by the graph above (for example, the change in democracy levels and state benevolence among the successor countries of the Soviet Union; this could be done, but I'm not up to it right now), no obvious associations jump out. A better test, perhaps, would look for changes in the degree of executive constraint (indeed, it looks as if changes in executive constraints do have a positive effect on a larger proportion of countries - 63% in my sample instead of 53%); but whether or not political regimes are associated with state malevolence and benevolence, other factors must be swamping much of their influence. Consider, for example, GDP per capita:

Fig. 13: GDP per capita (from the PWT) and the Physical Integrity Index, by regime type (as measured by the DD dataset)

As we might expect, in every regime type except for military dictatorships state benevolence is correlated with income per capita; the state is usually tamer in richer countries, though military dictatorships seem to get more malevolent the richer they are, even as they also become sparser as income increases. Or consider inequality (a much more striking picture):

Fig. 14: Inequality (measured using the UTIP data) and the Physical Integrity Index, by regime type (as measured by the DD dataset)

The malevolence of the state seems to be exquisitely sensitive to inequality in democracies, in contrast to non-democracies; the less repressive regimes are almost all on the upper left hand quadrant. This makes sense in light of the Acemoglu-Robinson story about the relationship between inequality and regime types: democracies enable class conflict, and hence the state is more likely to get more repressive as that conflict intensifies, whereas a dictatorship has "settled" such conflicts - arrived at some repressive equilibrium that is not especially sensitive to inequality. But of course other stories are also possible (not least that the data on inequality is not great); this is not a test of anything. It is also plausible to speculate that as the world became both more democratic and more unequal over the past 30 years, we would have seen a generally flat trend in the mean CIRI index; rising inequality would have cancelled out the effects of rising democracy.

Finally, consider economic growth:

Fig. 15: The Physical Integrity Index and per capita gdp growth, by regime type (as measured by the DD dataset) 

I confess that I found this picture surprising: I thought there would be an association between low levels of economic growth and greater repressiveness, but apparently not.

We could put this all together in some more complex model. But a structural break seems to remain; adjusting for gdp per capita does not change the picture in figure 11 much, though adding inequality softens the relationship a bit. At any rate, it seems as if the old idea of checks and balances is at least somewhat vindicated by the evidence of the last three decades: constraints matter, and don't count on benevolent autocrats.

(Code for all the graphs in this post is available here; as usual, it's very messy. You will also need this file of codes, plus the Penn World Table data, the CIRI dataset, the Political Terror Scale, the Polity data, the DD dataset, and the UTIP inequality dataset).

[Update 13/12/2012 - minor wording changes for the sake of clarity]

[Update: a quick graphical followup to this post here]