The origins and evolution of democracy: an exercise in history from a constitutional economics approach

Again, Tullock (1987a) has to be credited as the first author in having made this distinction. The point has been further elaborated by Wintrobe (1990, 1998), whose “tinpot” dictator would correspond to Olson’s bandit king. On the other hand, Wintrobe’s insights on the process of a modern dictatorship’s decline don’t include a model of its replacement by a democracy. Nor did Tullock (1987a, b).

Taken in its broadest sense, the term tax rate could include the risk of expropriation as well.

Bates (1987, pp. 41–42). Italics in the original.

Anthropologist Birdsell (1973) has shown that the tribal stage is subject to the same dynamics of continuous division as prevailing in the society of bands whenever a critical size of population is reached. With sustained increase in population, the repeated division of the tribe should eventually come to an end, just the same as that in the prior band society and due to the same reasons.

To give but two examples, see Service (1975) as well as the articles collected by Claessen and Skalnik (1978).

Since I resort to a graphical solution of the model, I voluntarily avoid any attempt to further specify the function of Y to t and G. For a half hearted attempt, see Mueller (2003, ch. 18).

The point has been raised before by Friedman (1977), Rogowski (1998) and by other authors as cited in the first section of this paper.

See Fleck and Hanssen (2002) on classical Greece.

On the relationship between the problem of control and the economic organization, see Alchian and Demsetz (1972). Turning again to medieval peasants, another condition for the effectiveness of the stick alone was their concentration for work, which facilitated the control of their effort. This seems to be an explanation of the manorial system and the lords’ preference for labour surplus vs. product surplus extraction more convincing than the ones proposed by North and Thomas (1973) and by their critics.

Friedman (1977) mentions another alternative: the combination of several states into one in order to increase migration costs for traders. He cites some historical cases in which this occurred, and his list could be extended to account for the imperial expansion of former state-cities. However, the success of these attempts is often subject to the hazards of war. Hence, I tend to think that the “parliamentarian” solution is more common, at least in the short run. Once a parliament has been created, the possibility of restricting the movement of even agricultural labourers may be somewhat hampered. This observation could help to answer the questions of why serfdom was not generalised again in Europe after the Black Death restored the scarcity of agricultural manpower, or why the agricultural labourers could move freely in the scarcely populated British colonies of Northern America or Australia. However, one should not exaggerate the extent of freedom in cases like these: one has to take into account also the high numbers of slaves, intended servants and coerced convicts living in the latter territories.

With precedents dating back to Hume and Kant, that relationship has been outlined by Jones (1988) and by Baechler et al. (1989). The more recent attempt I know of is the collective book edited by Bernholz and Vaubel (2004).

See Dudley (1990, 1991, 1997). This author highlights also the effect on taxes, though not on political representation.

A further effect of gunpowder on internal politics resides in that, the more territory in the hands of the King as a result of the working of the two preceding factors, the lower became the probability for a given city to be close to the border and, hence, the lesser its ability to transfer its loyalty to another monarch. This factor acts as a feedback which reinforced the process of territorial concentration.

As François put it gracefully: “My cousin [Charles V] and I agree in everything: we both want Milan”. Milan meant dominion on the route through the Alps, which was tantamount to European hegemony.

At least for an author, Latzko (1993), it is even dubious that we can speak at all about economies of scale in warfare prior to the invention of firearms.

Though Dudley seems to be right in classifying the Roman military technology as one leading to increasing returns to scale, the fact remains that the Romans often crushed armies that outnumbered them by far. There is no puzzle, however, if those enemies used an inferior technology, as it seems to have been the case.

This point has been stressed by several historians of warfare and civilisation, from whom has been taken by Dudley. See in English McNeill (1963, 1976). Dudley credits Altheim (1952) with the merit.

A paradoxical feature of Roman politics is the continuity and spread of (nearly) democratic forms of government at the municipal level. However, in so far as the Pax Romana greatly reduced the cost of migrations between cities inside the Empire, this is what we had to expect according to our proposition. Now then, the power of cities to oppose an increase of imperial taxes remained low, also in accord to it.

I explicitly renounce to offer here one more review on the question of whether the founding fathers of US Constitution reflected in it their own economic interests or not. The pioneering study on this subject is that of Beard (1913). The most recent contribution to my knowledge, including a thorough review to all prior work, is McGuire (2003).

As with the case of the autocrat, McGuire and Olson (1996), as well as Niskanen (2003) have preceded us in this path. Here, too, our explanation differs from theirs to some extent.

By applying the rule of the chain one can be obtained from the other. If we prefer, we can consider them as one, and resort to another condition that has to be met. Since G = Y · t, his condition can be stated as Y/G = 1/t. This has an immediate application to the graphical method that we have chosen to follow in this contribution.

See Schofield (2004) for a detailed account of fiscal innovation under the Tudors. Regarding the power of the English Parliament, acknowledging of the principle “Redress before Supply” meant a substantial difference with respect to those existing in other countries.

The role of the Parliament as the cashier of the Army and its further consequences has been stressed by Tullock (1987a, b). For this author, however, the ultimate causes of the power of the Parliament in the English case remain “mysterious” (p. 183). Other factors pertaining to our list seem to have been at work as well. The question of how to pay war debts was present at least in the 1688 episode, which, in other respects, resembled those of the 1640s.

He also mentions that the initial strategy of the American representatives in Britain, trying to create division between King and Parliament and among parties inside the latter, proved wrong, given the good harmony between the two key institutions that existed under George III: frustration of this seemingly costless strategy eventually led the Americans to choose the more costly way of war. As another factor, Reid contends that the groups more damaged by the British measures to some extent alleviated the cost of Revolution to the rest of the population by taking on their shoulders the task of government.

How exactly it was related is a question more difficult to elucidate. In our simplified model, low taxation provides an incentive to work and to invest. Complexities of real life had led historians and economists to highlight other ways of influence, as are: more secure property rights including intellectual property; less room for rent seeking, and enhanced independence of the judiciary. Even a cursory explanation of these mechanisms and reference to the authors that have proposed them would demand more room than is available in this contribution.

See, for an instance, Holcombe (2002). Public beneficence existed at the time, nonetheless.

See Downs (1957), as well as Meltzer and Richard (1981, 1983). Regarding the synchrony between general enfranchisement and extensive redistribution, this is most apparent in Europe. It seems likely that abundance of land and the resulting higher incomes per capita have delayed the distributional struggle in North America with respect to Europe.

The citizen under the veil of ignorance wouldn’t know whether she is bound to be rich or poor in the future. If only for this reason, and provided that the marginal utility of money is decreasing, she may feel adequate to redistribute some rent from rich to poor through the fiscal system. For a pure utilitarian, how much is to be distributed would be the outcome of a balance between, on one hand, the utility of the transfer in case of being a pauper multiplied by the expected probability of become impoverished, and, on the other, the disutility of paying the taxes necessary to sustain the paupers by the probability of being among those who have to pay. From the social welfare function perspective the optimum is identical.

This picture does not differ much from the predictions that can be made from the median voter models. Meltzer and Richard (1981) have shown that a considerable degree of vertical redistribution is possible even when the political decisions reflect the preferences of the median voter. The adverse consequences of this redistribution on growth could be big, but the outcome on social welfare wouldn’t be as adverse. Niskanen (2003) comes to more pessimistic predictions while attached to the median voter theory.

Again, McGuire and Olson (1996), as well as Niskanen (2003) have preceded us in the building of an equilibrium model of redistributive democracy. Again, however, our model differs somewhat from theirs.

For an extensive treatment, see Tullock (1989).

See, among others, Olson (1982, 1983), Choi (1983), Tang and Hedley (1998), Heckelman (2000), Coates and Heckelman (2003a, b), and Knack (2003).

Assessment of the relationships between public sector growth and the growth of income is plagued by serious econometric problems, though most contributions conclude that there is an inverse correlation. Among the immense literature published since the early 1980s, two papers that honestly acknowledge for these problems and attempt to solve them are De la Fuente (1997) and Fölster and Henrekson (1999). On the relationship between regulation and growth, see for instance Gwartney et al. (2004), Doucouliagos and Ulubasoglu (2006), Dawson (2006), and Weede (2006).

Practically limited to the contributions mentioned in footnote 32.

This is an implication of the communicable character of knowledge. The point has been extensively developed by growth economics and constitutes the cornerstone of the endogenous growth models.

This could be one of the reasons why the empirical literature that tries to measure the impact of distributional coalitions on economic growth hasn’t found big differences in the rates of growth of different countries, even when these appear to be subject to different pressures stemming from those coalitions (See references in footnote 34). Another reason could be that, as a result, the international differences in the extent to which governments pursue distributional goals aren’t so big after all. The same problem could affect the empirical literature that tries to assess the negative impact of public sector growth on overall economic growth (See references in footnote 35).

The number of contributions comparing the economic performance of democracies and dictatorships in our present World has increased quickly during the last years. Though the conclusion of this literature is mainly sceptical, this serves to illustrate the vulnerability of democracy to interest groups, as well as the fact that when an autocracy establishes in once democratic countries (usually with the applause of a part of the electorate), its behaviour may present important differences with that of more traditional autocracies, and resemble that of democracies in many respects… at least, as long as it allows for migration of people. As a sample, see Sirowy and Inkeles (1990), Przeworski and Limongi (1993), Farr et al. (1998), Wu and Davis (1999), Ali and Crain (2001, 2002), Tavares and Wacziarg (2001), Plümper and Martin (2003), and Mueller (2003, ch. 18).

See Tideman and Tullock (1976), Hylland and Zeckhauser (1979), and Tideman (1997).