r encouraging capital inflows until the domestic gross rate of return on capital is r’ = r.
This implies that w’ is also equal to w as given by (10.10) and therefore from the analysis above the coup threshold in the economy with capital inflows is given by ϕ* in (10.19). This establishes a version of Proposition 10.1.
Proposition 10.3: Consider the economic model and the democratic consolidation game described above and define ϕ* by (10.18) and ϕ* by (10.19):
• If ϕ < ϕ*, there are coups both before and after financial integration.
• If ϕ≥ ϕ*, there are no coups either before or after financial integration.
• If ϕ* ≤ϕ<ϕ*, there are coups before financial integration but not after.
Therefore, just like trade opening, financial integration makes democracy less redistributive. This implies that the elites have less to fear from democracy and are less willing to undertake a coup. In addition, with financial integration, factor prices again move toward world prices (i.e., returns to labor increase and those to capital decline), and coups again become more costly after financial integration. (Recall that before financial integration, coups also increase the return to capital and land but after financial integration they do not.) Both of these effects make democratic consolidation more likely after financial integration. The additional effect highlighted herein is that financial integration may also encourage the median voter in democracy to choose lower taxes to attract more capital and increase wages. This effect is discussed in greater detail in the next subsection.
Similarly, financial integration by making democracy less redistributive and the use of force against democracy more costly may help the transition to democracy. Therefore, as long as the choice for the elites is between democratization and repression, we could also state a proposition similar to Proposition 10.2; however, we refrain because the analogy is immediately apparent.
5.2 Capital-out
The previous subsection showed how financial integration can help democratic consolidation and the transition to democracy through a channel similar to the effect of increased international trade: by affecting the income gap between the elites and the citizens and by influencing the cost of using force against democracy. However, the more important role of financial integration may be the potential threat that capital may fly out and leave the country if taxed too heavily. To put this in context, imagine a Latin American country before financial integration. If capital is taxed heavily, it can withdraw into the informal sector or the elites may decide to consume more and save less. This is what we capture with our cost of taxation, C (τ). After financial integration, however, there is another option. If capital holders are taxed heavily, they can take their capital to Panama or the Cayman Islands, where taxes are lower. This increases the elasticity of capital with respect to taxes and affects how much taxation democracy would like to impose on the elites. In this subsection, we analyze the implications of this potential capital-out channel on the consolidation of democracy. To simplify the analysis and highlight the implications of the capital-out mechanism, we now abstract from capital inflows; therefore, after financial integration, factor prices do not change.
Assume that we start with the economic model described previously and there is no trade in goods or financial flows, so factor prices are given by (9.2).The rate of return to capital is now θ ((K + σ L )/(1 - δ))θ-1 and with the tax rate, τ, the net return is:
If capital flies out, it has a (net) rate of return r. We assume that:
Therefore, if the citizens set their unconstrained tax rate, τp, capital will fly out. As long as capital is sufficiently important in the income of the elites and therefore in the tax revenues that the citizens collect from the elites, the citizens would not want to tax incomes at such a high rate that capital holders take their money outside the country. Therefore, in equilibrium, they have to set the lower tax rate,p, such that:
(10.29)
Given this lower tax rate, democracy becomes less costly and more likely to be consolidated. Similarly, it also is less attractive for the elites to use repression to avoid having to democratize.
More formally, after financial integration, the returns from democracy are now given by:
(10.30)
which are simply (9.7) evaluated at the tax rate,P. We again use the notation V to refer to values in the open economy. Because the tax rate that applies after financial integration,p as given by (10.29), is lower than that which applies before financial integration,p as given by (4.11), we have that:
That is, democracy is better for the elites after financial integration.
Similarly, imagine the values of democracy to the citizens and elites when the citizens promise to set a tax rate of zero, with this promise upheld with probability p. From (9.15), these are now given by:
These expressions, once again, take into account that if democracy gets to reset taxes, the median voter sets the lower tax,p, instead of τ pbecause at the higher tax, τ p, capital will fly out of the country.
Even after financial integration, the costs of coups are not different because after a coup, there is no taxation and, therefore, no capital flight. As a result, the values after a coup are still given by (9.24). We again define ϕ* by (10. 18) as the threshold value before financial integration (i.e., at ϕ = ϕ*); we have Vr( D, τD = 0) = Vr (C, ϕ). Also define ϕ* as the corresponding threshold after financial integration - that is, such that at ϕ = ϕ*, we have:
which implies that:
(10.31)
That democracy is less redistributive after financial integration immediately implies that:
Consequently, in the current model, Proposition 10.3 again applies but because of the effects of financial integration through the potential of capital flight rather than through capital inflow. Therefore, financial integration again may help democratic consolidation but now through as somewhat different channel. After financial integration, democracy does not find it optimal to impose as high taxes because such taxes would induce capital holders to take their assets abroad. Because democracy is expected to be less redistributive, the elites do not have as much to gain from a coup and democracy is more likely to survive even during periods of crisis.
A similar argument also applies to transitions to democracy. Consider the transition to democracy game discussed in the previous subsection and recall that the values to the citizens and the elites from repression are still given by (10.21). Make the same assumptions as in the previous subsection so that concessions by nondemocracy do not work; after financial integration, the elites prefer democratization to repression if:
where Vr(D) is defined by (10.30). This condition defines a different cutoff level, now denoted by K*, such that for all ≥ K*, the elites prefer democratization to repression after financial integration. More explicitly, K* is given by:
(10.32)
which is simply (10.24) with τP replaced by τp. ThatP < τP immediately implies:
where K* is given by (10.24). Because the presence of financial integration makes democracy less bad for the elites, repression has to be cheaper for it to be optimal.
As long as the choice for the elites is between democratization and repression, this analysis leads to a proposition paralleling Proposition 10.2, where now financial integration - by again making democracy less redistributive - may lead to democratization in circumstances in which without financial integration the elites would have preferred repression. We do not state this proposition because its logic is clear and its implications are identical to the results already stated herein.
6. Increased Political Integration
Another dimension of globalization is increased political integration. In a more globalized world, there are closer political links between nations. A common view is that the increased integration of Eastern European nations with the European Community has been an important element in their smooth transition to democracy and in the rapid consolidation of their democracies. Supporting this view is a
finding that the post-Communist societies that are geographically closer to Western Europe (e.g., closer to Berlin) are more democratic (Kopstein and Reilly 2000).
A natural reason for this is that through various channels, greater political integration between democratic and nondemocratic societies increases the costs of using force to prevent democracy. The reasons might vary, ranging from potential sanctions or boycotts by democratic nations if there is a coup against democracy, to the destruction of trading relationships. Another complementary channel would be that with greater political integration, civil society in nondemocratic nations or in unconsolidated democracies becomes stronger, increasing the cost of coups or repression.
In a reduced-form way, we capture all of these ideas by supposing that the cost of using force against democracy — more specifically, the cost of coups - increases after political integration. In particular, assume that after political integration, a coup leads to the destruction of a fraction> ϕ of the asset stock of the elites, whereas before political integration, the same fraction was ϕ. This implies that after political integration, the values to the citizens and the elites following a coup are:
(10.33)
instead of (9.24). We can now state the following proposition:
Proposition 10.4: Consider the models underlying Propositions 10.1 and 10.3. Once political integration takes place, the cost of a coup is higher, and a society is more likely to be a consolidated democracy.
This proposition, therefore, shows how political integration may have effects similar to those induced by increased international trade and financial integration. However, although these effects ultimately have similar implications, they work through a different channel. They discourage coups by making them more costly because of international pressure and sanctions or because of the induced development in civil society that follows from closer political links among democratic and nondemocratic nations.
7. Alternative Assumptions about the Nature of International Trade
Propositions 10.1 and 10.2 illustrate that globalization in the form of increasing integration of markets for goods may promote both the creation and consolidation of democracy. These results, however, stem from the structure of the models we wrote down. Most models of international trade have the implication that trade promotes the income of the poor in developing countries (who we are associating with the citizens) because such countries are typically labor-abundant. However, as already alluded to, the empirical evidence is somewhat equivocal about whether increased trade in fact promotes equality in poor countries. Also, it could be that some developing countries, such as Argentina or Chile, are in fact land-abundant and not labor-abundant. In this case, increased globalization has the effect of increasing the rate of return on land. This not only increases inequality but it also raises the incomes of the asset holders who, as we argued extensively in Chapter 9, have most to lose from democracy.
Here, we sketch a different type of trade model, motivated by a salient empirical pattern in the recent data: in many of the less developed nations opening to trade during the past twenty five years, returns to skills and, therefore, income inequality, have actually increased (Leamer 1995, 1998; Cragg and Epelbaum 1996; Attanasio, Goldberg, and Pavcnik 2004). This is the opposite of the prediction of the simple Heckscher - Ohlin trade model because the less developed nations in question are relatively scarce in skilled workers. So, everything else being equal, trade opening should reduce the skill premium in those countries. A lengthy discussion of why returns to skill appear to have increased in these countries is beyond the scope of this book, but there are some natural conjectures. Most important, there is wide consensus that many of the important advances in technology during the past twenty five years have been relatively “skill-biased,” meaning that they favored skilled workers and, everything else being equal, tended to increase the skill premium (e.g., Acemoglu 2002). Most of these technologies are embedded in machines produced in the United States and in some OECD economies. Less developed countries can make use of these technologies only if they import the relevant machines from the United States, and other rich nations. This implies that when they are closed to international trade, less developed nations typically do not use these machines. The important implication for this analysis is that trade opening now comes with a change in the technology of production toward more skill-biased technology, increasing returns to skill. So, there will be a technology effect counteracting and perhaps dominating the standard Heckscher - Ohlin effect of trade reducing the skill premium in less developed nations (Acemoglu 2003b; Thoenig and Verdier 2003).
What are the implications of this for democracy? The discussion so far, which was based on our two-class model, might suggest that implications of the spread of skill-biased technology to less developed nations is the opposite of what we emphasized: an increase in inequality and, therefore, a force acting against the creation and consolidation of democracy. This is certainly one possible conclusion; however, other potential forces may be at work. The increase in the returns to skills and, more generally, the increased role of human capital in the modern economy (discussed in Chapter 9) can be interestingly analyzed in our three-class model. In the context of that model, we can think of skilled workers as constituting the backbone of the middle class. This implies that trade opening, associated with the transfer of skill-biased technology, increases the income of the middle class. As discussed in Chapter 8, the middle class can act as an important buffer between the rich and the poor, and an increase in the income of the middle class may help the creation and consolidation of democracy. This is because the median voter in democracy may be a member of the middle class, and an increase in the income of the median voter (relative to mean income) reduces the propensity of democracy to be anti-rich. With less radical policies adopted in democracy, the rich have less to fear from democracy and are less willing to use force to prevent democratization or create a switch to nondemocracy.
To highlight these issues, we briefly return to our basic model from Chapter 8. In that model there are three groups of agents; the rich of size δr, the middle class of size δm, and the poor of size δp. We normalize total population to 1 as before; thus, Σiδi = 1. We assume that δP > δm > δr; that is, the poor are the most numerous, then the middle class, and the rich comprise the smallest group in the population. Also, we denote average income byas before and let incomes be given as in (8.1 ) where (8.2) holds so that the rich are richer than the middle class, who are in turn richer than the poor.
To simplify the analysis along the lines of the discussion in Chapter 8, we assume that δP < ½ , so that the poor do not constitute an absolute majority and a middle-class agent is the median voter. Moreover, suppose that θm< δm or> γm, so that the middle class is less rich than mean income and would like to impose some amount of redistributive taxation. As in Chapter 8, the amount of redistribution preferred by the median voter - therefore, that which results in an unconstrained democracy — is given by the tax rate τm, which satisfies (8.5).
Now consider the basic political game discussed in Chapter 8 in which the society is nondemocratic with the poor and the middle class excluded from voting. Because of a potential revolution threat, the rich are considering democratization or the use of force (repression) - the promise of limited redistribution is not credible. As in the previous analysis, there is a cutoff level for the cost of repression,(τD), given by (8.29). When the cost of repression, K, is equal to(τD), the rich are indifferent between democratization and the use of force. They prefer repression whenever K < (τD). It is clear from (8.5) that an increase in the income share of the middle class reduces τm and, therefore, decreases(τD).
Opening to international trade and the associated transfer of skill-biased technology by increasing the incomes of the middle class may reduce redistribution in democracy and help induce a transition to democracy. In particular, suppose that after trade opening, because technology becomes more skill-biased, θm increases tom
and, as a result, the most preferred tax rate by the middle class falls tom given from (8.5) implicitly by (δm — m)/δm = C’(m). This implies that the new threshold for the rich to be indifferent between repression and democratization becomes:
(10.34)
wherer is the share of the rich in incomes after trade andr < θr because θm > θm. Clearly, we have
Then, we can state the following:
Proposition 10.5: Consider the transition to democracy game described above. The society starts nondemocratic with the poor and the middle class excluded from voting. Define (τD) by (8.29) and , which applies after opening to international trade and the transfer of skill-biased technologies by (10.34). Then, we have that:
• If K < , the rich use repression to prevent a revolution and democratization both before and after trade opening and transfer of technology.
• If K ≥ (τD), there is democratization both before and after trade opening and transfer of technology.
• If ≤ K ≤ κ < (τD), the rich use repression to prevent a revolution and democratization before trade openingand transfer of technology, but there is democratization after trade opening and transfer of technology.
Economic Origins of Dictatorship and Democracy Page 48