Vol.39(1-2), March-April 1991, pp.22-34.


George Modelski & Gardner Perry III

Department of Political Science

University of Washington


Paper presented at the International Conference on "Diffusion of Technologies and Social Behaviour: Theories, Case Studies, and Policy Applications" held at the International Institute for Applied Systems Analysis (IIASA), Laxenburg, Austria, June 14-1 6, 1989.



We wish to thank Ted Gurr and the Center for Comparative Politics at the University of Colorado for making available data from the Polity II dataset, and Raymond Gastil and James Huntley for their comments.




Democracy is a mechanism of collective choice and a form of social organization that can be considered a superior substitute for other such mechanisms or forms of organization. As such, democracy may be expected to grow, or diffuse, over time, amongst the world's population, and the question posed in the present study is: does that growth follow a regular pattern, according with the Fischer-Pry substitution model of technological change?

Our inquiry shows that, prior to 1800, democratic development was experimental in character, but it has been growing fairly rapidly since the middle of the 19th century, generally fitting quite closely the model of diffusion. At the present time, about forty per cent of the world's people live in democracies; by extrapolation the model suggests that the democratic community might reach the 90 per cent level toward 2100.


Democracy and democratization


Democracy literally means rule of the people. More analytically speaking it might be seen both as a technology of collective choice, and also as a type of social behavior or community organization. The technical, or instrumental, aspects of democracy are best recognized in the election process, a form of macrodecision by which a community selects some individuals to positions of public leadership. In this sense, practical democracy consists of a number of rule-governed devices - media, parties, voting, and majority rule - and of policies executed via working mechanisms of representation. The substantive, or expressive, aspects of democracy, on the other hand, relate to its characteristics of equality and freedom, and they reflect the optimum conditions in which democracy operates effectively: a community of equals (symbolized by "one man, one vote") in which, in a state of autonomy, all live as free individuals. The first recorded discussion of democracy aptly combines these two aspects in the phrase "equality under law" [1].

Democratization is the process of building, or creating, democracy. That process, too, moves along two distinguishable, if related, paths: wherever democratic techniques of macrodecision are discovered and spread, a process of diffusion of these innovations occurs. Alternatively, we see it also as the process by which democratic communities grow, via a form of clustering (or concentration) into larger communities of democracy; for the evolution of new types of community, too, is a form of innovation. The two processes are obviously interdependent: the diffusion of democratic procedures will produce no more than `formal' democracy unless rooted in, and nourished by, conditions of greater equality and freedom, where procedural rights are effectively exercised and are seen to work; a society of equality and freedom cannot last without observing democratic procedures.

Viewed from another angle, democratization may be either intensive, or extensive. By intensive (or vertical) democratization we understand the change in the quality of the democratic experience in a given community; that is, we ask: how good is this particular democracy? and we may wish to judge it e.g. by the extent of voting rights, the working of representative institutions, or the presence of democracy at the several levels or fields of social organization (such as national, local, political, economic, etc). We might also want to judge it by some general standard of "perfect democracy" (analogous to "perfect competition" in economics). That means that nation-states, and communities, might be seen as less, or more, democratic, and that the quality of democracy is subject to change over time, in both directions, positive and negative. Conceptions as to what is democratic might also change, and the point at which a given community is be called democratic might therefore be a variable one.

Extensive (or horizontal) democratization, on the other hand, measures the quantitative extension of democratic communities, and their global spatial reach. It asks: how much democracy is there in the world? The present paper emphasises this latter question.

The process of democratization occurs in an "environment" (or niche) of finite "carrying capacity". We conceive of it as unfolding within the limits of the world system, in the context of its evolution, and as part of it. That is, democratic procedu res are diffused inside a `market' that extends to, and is limited by, the social organization of the human race on this planet, and democratic communities grow as part of world system evolution. We assume that democracy is not limited in its reach to certain regions or cultural areas or special circumstances. Hence the limit of the process of democratization, its potential, is the world population but we also need to bear in mind that that population has been a steadily expanding one, by more than one order of magnitude in the period under discussion.

As a set of techniques, and principles of community organization, democracy is in some respect a superior substitute for other technologies, and forms of community organization. In its classic Greek form, democracy was the basic alternative to monarchy, with aristocracy an intermediate solution. In the modern era republics and (increasingly absolute) monarchies emerged as the two principal forms assumed by the rising nation-state. Gradually, republics (and constitutional monarchies) evolved into democracies as substitutes for absolute rule, or for narrowly-based, arbitrary, despotic, or dictatorial forms of organization, and they have been shown to have wide potential appeal, on such grounds as flexibility, accommodation of variety, self-legitimating properties, and capacity to civilize conflict. On such a view democratization has been a process of gradual substitution of democratic for non-democratic forms of organization. Or else as a persistent problem of choice between alternative and competing forms available to communities for dealing with their problems.

Democratization is, moreover, a time-bound process. We do not expect it to be completed, all at once, world-wide, but rather to spread in stages, gradually, and slowly at first. We would expect it to emerge following a series of experiments, some of which are bound to falter, and only a few to succeed. We know that the first trials with direct democracy occurred in relatively small urban communities in ancient Greece, prominently in Athens, ca. 5OO - 3OO BC. They did not last, but their failure was a noble one and is remembered to this day. Another series of attempts at popular rule materialized in Italy, after the year 1OOO, among self-governing cities of which Venice was the most successful, but there, too, most experiments deteriorated into tyrannies, and even Venice ultimately turned into an oligarchy.

The seeds of modern democracy began to sprout after 15OO as alternatives to unfettered monarchy which was then prevailing as the answer to the question of the character of the modern nation-state: i.e. was the state to have unlimited or limited powers? [2] The critical events were the wars of religion in Western Europe after 1561: they produced both the absolute monarchies of Spain and France, and the republican United Provinces of the Netherlands. In limited monarchies, this republican "strain" went under the banner of "commonwealth" (as in John Locke's "Two treatises of government). Helped though it was by the Venetian model, now transposed to the larger scale of nation-states, the republican way advanced sl owly at first. Eventually, it acqui red a strong liberal flavour and came to be strongly tied (as in the British case) to representative institutions.

Democratization took a leap forward in the l9th century in the form of representative democracy. But it could not have done so without suitable techniques of representation that had been previously devised in the republican and liberal states. That is why the achievements of the 16th-18th centuries are an essential part of the master trend of democratization.

The propensity to view democracy in long perspective can hardly be claimed as altogether new. Having observed the American experience with Jacksonian democracy, Alexis de Tocqueville (1969:9) announced more than a century and a half ago that "a great democratic revolution ... is taking place in our midst" and added that some saw it as "irresistible" because it is "the most continuous, the most ancient, and the most permanent tendency known to history".[3] But this revolution, if we are to credit deTocqueville's term, is still in progress, and its precise shape remains to be determined.


Democracy as innovation


In long perspective, democratization can therefore be regarded as a process of structural change, or innovation or, better still, as a cluster of innovations. An innovation is an idea or practice that is new to an individual or a community. Democracy certainly was such an idea when its techniques originally evolved in a simple but persuasive form; it was again innovative in its modern version, when it had to be adapted to a larger scale of social organization and (by continuous evolution) has come to form a repertoire of some complex new practices. While in the "old democracies" this inventive aspect is apt to be forgotten, it is much in evidence in those communities where democratic practices are still unfamiliar.

The question arises: is it not likely that democratization, as a process of innovation diffusion, is in fact also a learning process?

We know that innovations spread in a diffusion process involving regular patterns in social systems over time. According to Everett Rogers (1962:67) "the process by which innovations are adopted by individuals is essentially a limited example of how any type of learning takes place". More generally, we might assume with Cesare Marchetti (198O:267) that "society is a learning system", that learning is "random search with filters" that brings about lasting changes in behavior, and that such "random searches" are characterized by logistics (a curve representing a function involving an exponential and shaped like the letter S). Logistic functions have been used to study the growth of human and other populations.

"A general finding of past investigations" Rogers argues is that the pattern characterising the way innovations are adopted (that is, "adopter distributions") follows " a bell-shaped curve over time and approaches normality ... this type of distribution is essentially S-shaped when plotted on a cumulative basis".

Three reasons have been adduced why adopter distributions might be expected to approach normality: early sociologists, including Stuart Chapin, who studied the city manager form of government in the United States, observed that the adoption of an idea followed a bell-shaped pattern. The learning of individuals follows a normal curve and the gain in learning per trial is proportional to the product of the amount already learned and the amount remaining to be learned before the limit of learning is reached. As the individual learning curve is extended to the case of a social system, experience with the innovation is gained as each successive member adopts it. Last, the interaction effect is the process through which adopters of an innovation influence those members who have not yet adopted; the group pressure for adoption rises as the number of adopters increases (Rogers 1962:152-155).

This time pattern of diffusion in turn suggests a categorization of adopters. The simplest might be one that distinguishes between innovators, early adopters, late adopters, and laggards. We might regard as innovators the first 10 per cent of adopters, and the next forty per cent as early adopters. At this point (50 per cent, the population mean, or the flex point) adoption ceases to increase at an increasing rate, brings in the "late adopters", and ultimately levels off for the "laggards" (the last ten per cent).

In this paper we shall apply one basic form of the diffusion model, the Fisher-Pry (1971) model of technological substitution, on the assumption that democratization may be considered as a competitive substitute for other methods of social organization. Basic human needs for organization and collective choice need to be satisfied, and democracy provides that better substitute. The Fisher-Pry model expresses that process, and its most conv enient form is (1971:77):


F/(1-F) = exp 2 alpha (t-t ),


where F represents the fraction of substitution (in our case, the "fraction democratic"), and 1 stands for the "size of the market" (in our case, the world population). Then the slope of the line is 2 alpha, and t (in years) is the time constant (calle d by Fisher and Pry the takeover time), defined as the number of years required to go from 10 % to 90 % saturation level. The mid-point of that range (50 %) is the flex point, t . Then a plot on semilog paper of F/(1-F) as a function of time allows one to fit a straight line through the data and make appropriate extrapolations.

In the remaining portion of this paper we shall ask: Can democratization be shown to be a major pattern of the world system? Might it be no more than a series of unrelated national developments governed by local conditions? If not, then how do we document it and what is the evidence for it? What has been the shape of (horizontal) democratization over time?


This is a large field; for our present brief, we propose the following course:


1. to operationalize the question by asking: what proportion of the world population has lived in democracies at various times since about 15OO? and

2. to further ask: does this information fit the substitution (or learning) model and if so, what might be the implications?

For if the cumulative growth of the share of world population living in democratic communities (that is, in the global democratic community) can be described by a logistic-type (S-shaped) distribution (that in a semi-logarithmic plane appears as a straight line) then we have grounds for arguing that democratization is a social learning process (of innovation-diffusion/community growth).


The data

We propose to test our hypothesis against two sets of data now available on the population of democratic communities since about 15OO.


Our data on these communities is in two parts:

1. For the period 1800-1986 our basic source is the POLITY II dataset (Gurr, Jaggers & Moore 1989); the background to that collection is explained in Gurr, Jaggers, & Moore (forthcoming).

The Polity II survey covers "all independent members of the international system", those that have attained independence by 1975 and whose population exceeded one million by the mid-1980s. [4] It gives, for each such polity, an annual score of "institutional democracy", on a scale ranging from zero to ten. The score is constructed from three codings, those for (a) competitiveness of political participation - via the party system; (b) openness and competitiveness of political recruitment, that is quality of the electoral system, and (c) constraints on the chief executive (checks and balances). There are no coded data on human rights or political liberties.

The "institutional democracy" score is constructed additively from these three indicators, so that party competitiveness rates a maximum weight of three, executive recruitment via elections also a maximum of three, and restraints on the chief execut ive, four. It measures the degree of intensity of democracy for each polity.

Selected for inclusion in this survey since 1800 are all polities that had, in any year since 1800, an "institutional democracy" score of six and above. There was one in 1800 (United States, with a score of seven); fourteen in 1900, and 46 in 1986 ( of which 29 scored ten). The most opulous national polities and their Polity II scores in 1986 are shown in Table 1.
(Click here for Table 1)

(2) For the period between 145O and 18OO we apply the same criteria (parties, elections, checks and balances) and the same scoring weights and ask: what are the polities that can reasonably be described, on these criteria, as "democratic experiments": that is those exploring a variety of paths of trial and error, at first in the republican, and later also in the liberal traditions, all generally tending in the democratic direction.

This yields a tentative list of eight independent polities as follows [5]: Venice (145O-164O); Swiss Confederation (1499-1531) [6]; Poland (15O5-1605) (Poland-Lithuania after 1569); the Dutch Republic (1579-1787); Britain (1688-18OO); Sweden (1718- 1771); United States (1776-18OO); France (1789-1799). Each rates a score of at least four on the Polity II scale; a somewhat lower threshold is used to take account of the experimental character of these early cases. Only two (Britain, the United States) remain in 18OO which is a time of crisis amidst global war, and only the United States carries over to the Polity II survey, where it rates a seven. Britain keeps a score of four until 1837.

The data on democratic communities are then combined with population figures, for the relevant communities, and for the world, based on McEvedy and Jones (1978) and other sources cited in Perry (1987). The populations are those of the metropolitan territories, and do not include dependencies.


The analysis

Table 2: World Population Fractions summarises in Part (a) the data for the period 1450-1800, basically the era of the innovators. The period up to about 1750 is one of experimentation, with a world share of population ("fraction democratic") ranging between one and two per cent. The same data is presented graphically in Figure 1: Experimental Democracy 1450-1800 [7].

Table 2, Part (b), summarises the data for the period 1800-1986, the era of the "early adopters". The same data is presented graphically in Figure 2: Institutional Democracy 1800-2000. But starting the analysis in 1837, as in Figure 3: Institutional Democracy 1825-2125 yields a trend line with a somewhat lower growth rate [8].

The American and the French Revolutions launch after 1776, the stage of early adoption, even though the American Revolution is, in the first place, seen mainly as a republican achievement, and the French Revolution soon loses its democratic character. But the democratic trend takes off strongly in mid-19th century, moves past the ten per cent range and in the late 2Oth approaches the flex point. This theoretical half-way mark, when 5O per cent of the "market" could be expected to be saturated with democratic practices, is, for Fig.3, the year 2003; the actual "fraction democratic" approaching the 40 per cent level in the late 1980s.

The overall directionality of the process is unmistakable. But the process is not unilinear; both peaks and setbacks are strewn along the way. The early record, the (a) series, shows two peaks, in 16OO, and in 179O, each followed by declines, when the proportion fell as compared with the preceding decades. In the more detailed, and more recent, series (b), we observe four abrupt increases; a major spurt in 1837 (when Britain enters), and going on to 1877 (France), 1918-21 (post-World War I) and 1945 -50 (post-World War II). A perceptible decline marks the years after 193O, especially World War II. In other words, democratization shows fluctuations in a generally upward movement; the fluctuations are due to "lumpine ss" (when large countries become - or cease to be - democratic); the process is also adversely affected by global wars.




Our test has been conducted with two sets of data, but with the same classification and scoring. Should other data become available, using, for example , somewhat different criteria of democracy, such as those incorporated in the annual surveys conducted by Freedom House after 1972, (see Gastil 1989a,1989b) and that also give additional weight to civil or political liberties then the test must obviously be replicated. It appears though that these two particular surveys are not that dissimilar in their rankings. For the year 1986, the difference between the Polity II and the Freedom House surveys, expressed in "fractions democratic", as shown in Table 3, is only some 2.4 points.

Our analysis shows that the cumulative growth in the fraction of the world's population that comprises its democratic communities conforms to the Fischer-Pry model. Is this no more than a descriptive finding? How might we explain this unexpected regularity?

For such an explanation (as already indicated), we reach to theories dealing with innovation diffusion, and learning, phenomena that in societal dimensions and in long perspective might best be viewed as collective evolutionary processes. Democratization is a pervasive process of structural change at the global level that effects a host of other developments and serves, indeed, as an envelope curve for most of what we commonly view as technological and economic innovation. It is also a long-range process, with a time constant on the order of 200 years. If the trend line for "institutional democracy 1800-1986" is extrapolated into the future, the world will be passing the 9O percent mark in 2075; if we take as our guide the analysis that treats the current pulse of democratization as having begun in mid-19th century, and we extrapolate our data starting with 1837, then the 90 per cent level is attained in 2117 (as shown in Figure 3). Either way the status of democratization as a major evolutionary process is unmistakable.

The process has, in the first place, the aspect of the diffusion of a technology of collective choice, and can be easily understood as such. The wide dissemination of improved artifacts of life (or "better mousetraps") is, after all, one of the most obvious facets of civilization. It can be studied, for example, by asking who are the experimenters and the innovators, and how they are followed by early adopters, and come to be engulfed in the great majority, or else by establishing the mechanisms and conditions facilitating or obstructing diffusion: proximity, similarity, or opportunity for interaction.

The process might also be viewed as one of community formation: one by which cooperation gradually evolves into more stable and institutional forms, first by trial and error experimentation, and then by a species of clustering (or nucleation), that via coalitioning, branches out into structures of global cooperation. This second aspect, about which we know less than about diffusion proper, makes it clear, though, that the process must be a slow one because communities change only at a deliberate pace and grow but gradually.

What lends stability to this process, and credibility to our extrapolations? Might it not be said that the Fischer-Pry model assumes an unchangeable environment, whereas changes are bound to affect future structural change in unfathomable ways?

Democratization (which might be called the Kantian process [9]) is one of a family of global collective evolutionary processes. Other members of that family are the Kondratieff wave, and the long cycle of global politics. The Kondratieff, with a time constant of over 50 years, governs global economic innovation and exchange. The long cycle, with a period of just over 100 years, and centered on the roles of global leadership and challenge, shapes the structure of global politics (Modelski 1987, 1990). These are coupled processes that are mutually reinforcing, jointly ensuring a dynamically stable development.

The role of global politics in all this is essential. The mechanisms of diffusion and clustering are set in motion by the long cycle. In that process the role of global leadership has served as source of innovation, and via demostration effects, as center of innovation diffusion, as well as the nucleus of the emerging global democratic community. From the success of nation-states performing that role, the Dutch Republic, Britain, and the United States, has sprung the process of reinforcement that is essential to learning (the key to which is the proposition that "reinforced behaviour becomes more frequent"). The world powers have been the preferred models of imitation, and also, successively, the centers of gravity for community organization. [10] It is precisely when these centers at times ceased to "hold" (as in the 193Os, or about 18OO, in the 167Os or the 158Os) that the prospects of democratic organization dimmed. The long cycle of global politics, itself a learning process, intermeshes with, supports, and in turn derives strength from, the evolution of the global democratic community (Modelski 1987,1989,199O).

Viewing democratization in long perspective has enabled us to see it as the slow and gradual substitution of republican, liberal, and democratic regimes, for monarchical, absolutist, and dictatorial (or autocratic) forms of government and society. The question might be asked: do we expect this process to reach some form of equilibrium between democratic and dictatorial (or autocratic) forms, or should we look forward to the steady expansion of democracy through its entire global "market", that is through the bulk of the world's population?

Our view inclines toward the latter interpretation, and significantly on grounds having to do with the present phase of global politics that calls for world-wide "socialization", or "civilizing" of conflict. If global nuclear war is indeed unthinkable, then alternatives to it need to be put in place. and the obvious forms of such socialization, or civilizing, are the forms that have been nurtured in the emerging democratic community. Those which democracy has been substituting for, forms of absolute and dictatorial rule, have been too closely linked to the origins of the global wars of the past and do not offer a sound basis for development, or for dealing with other global problems, including environmental ones. It is from within the democratic community that substitutes to global war as a mechanism of collective choice are likely to emerge, and it is from within it that new forms will materialize that will be the substance of competitive substitution in the future.



1. In Herodotus' The Histories (Book III, 83) (written ca.450-430 BC), where it is compared and contrasted with monarchy and aristocracy. Writing a century later (in Politics III,7), Aristotle added new distinctions to these terms but, like Plato, he also put a negative spin on the term democracy.

2. Early modern writers on politics were familiar with the concept of democracy from classical sources; prominent examples include Niccolo Machiavelli (in the Discourses, printed in 1531), Gasparo Contarini (whose book on the Venetian constitution originally published in 1543 quickly became a classic of republican and constitutional thought), Andrzej Frycz Modrzewski 1551), Hugo Grotius (1600), Thomas Hobbes (1651), John Locke (1689), Jean-Jacques Rousseau (1762), James Madison (1787), and Immanuel Kant (1795). But they all saw it as direct democracy and doubted its practicality, except perhaps for small states (Rousseau), and might have been influenced by the negative spin put on it by Plato and Aristotle; in a republic, on the other hand, they saw at work the principle of representation, together with the idea of "commonwealth" (a whole body of people united by common consent to form a political community, a res publica). For Madison a republic was "a government in which the scheme of representation takes place" and it was to "modern Europe" that the "great principle of representation" was owed (The Federalist" Nos. X, XIV). Not until the publication of Democracy in America" in 1835 did the concept of democracy regain wide acceptance, and this time, in a representative form.

3. Woodrow Wilson (1918:35) wrote that "democracy seems about universally to prevail. Ever since the rise of popular education in the last century ... the spread of democratic institutions...promise(s) to reduce politics to a single form... by reducing all forms of government to Democracy". When James Bryce (1921:24,42) posed the same question: "whether the trend toward democracy ... is a natural trend due to a general law of social progress" his answer was less sanguine: "although democracy has spread...we are not yet entitled to hold...that it is the natural and therefore... the inevitable form of government".

4. This means that a number of small communities (such as Barbados, Belize, Kiribati, Malta, Solomon Islands, etc) have thus been omitted, many of which are democratic.

5. Cf. Cole's (1987:88) table of "Liberal Regimes in the World System since 1600" to which two earlier cases have been added: Venice (factions, elective chief executive, checks on chief executive by Grand Council), and Poland-Lithuania (factions, elective chief executive, checks on chief executive by Diet; described in Reddaway et al. 1952:440 as "Gentry Democracy"). Portugal ca. 1500 might be a border-line case (proto-parties, limits on chief executive), as well as Florence (1494-1512). This listing is entirely tentative and is meant to highlight representative trends at the national level of organization rather than draw up an exhaustive and definitive inventory of free and self-governing communities.

6. The Swiss Confederation, emerging out of the Everlasting League of 1291, became virtually independent from the Empire in 1499. By 1513 it consisted of 13 cantons, the government of some of which, including Schwyz, Uri, and Unterwalden, was a form of direct democracy (by an assembly of all male citizens of full age). By 1531 (war of Catholic cantons against Zurich) religious conflict divided it and "common action became impossible" (Langer 1972) yet this did not exhaust the persuasive power of the Swiss example. Such models were a source of inspiration i.a. for political writers in the Dutch republic ca. 1600.

7. The equation for the trend line in Fig.1 is y = .009x - 19.6; R squared is 0.7.

8. The equation for the trend line in Figure 2 is y = .025x - 49.6; R squared is 0.915. The time constant delta t (10-90 %) is 176 years; the flexpoint (50 per cent) is 1987, and 90 per cent saturation is reached in 2075. Calculating a sever-year moving average raises R squared only slightly, to 0.919. For the years 1825-2125 (Figure 3; 1837-1986 data), the regression line is y= .019x - 38.54, and the time constant is 228 years. The ten per cent level was reached in 1889; the flex point is 2003 and 90 per cent saturation is reached in 2117; R-squared is 0.893.

9. Writing in 1795, Immanuel Kant proposed the hypothesis of a self-organizing social process tending to bring about the condition of perpetual peace via i.a. the creation of republican regimes.

10. For a study of the post-1945 growth of the "Atlantic-Pacific" system as a basis for a community of democracies see Huntley (1980).


Anderson, Perry (1974) Lineages of the Absolutist State, London: NLB.Bryce, James (1921) Modern Democracies, London: Macmillan.

Cole, Timothy (1987) United States Leadership and the Liberal Community of States, Ph.D. Dissertation, University of Washington.

Contarini, Gasper (1599) The Commonwealth and Government of Venice, London: Edmund Mattes.

Dahl, Robert A. (1989) Democracy and its Critics, New Haven: Yale University Press.

Fisher, J.C. and R.H. Pry (1971) "A Simple Substitution Model of Technological Change" Technological Forecasting and Social Change Vol.3, 75-88.

Gillespie, John V. & Betty A. Nesvold eds. (1971) Macro- Quantitative Analysis: Conflict, Development, and Democratization, Beverly Hills, CA: Sage.

Gastil, Raymond (1989a) Freedom in the World: Political Rights and Civil Liberties 1988-1989, New York: Freedom House.

----- (1989b) "The Future of Democratic Progress and the Prospect for Peace", paper presented to the Pacific Northwest Colloquium on International Security, Seattle, May.

Gurr, Ted Robert, Keith Jaggers & Will H. Moore (1989) POLITY II CODEBOOK, Center for Comparative Politics, University of Colorado; April 7.

----- (forthcoming) "The transformation of the Western state: the growth of autocracy, democracy, and state power since 1800" Studies in Comparative International Development.

Huntley, James R. (1980) Uniting the Democracies, New York: New York University Press.

Langer, W. et al. (1972) The encyclopedia of world history, Boston: Houghton Mifflin.

Marchetti, Cesare (1980) "Society as a Learning System" Technological Forecasting and Social Change, Vol.18, December, pp.267-282.

Modelski, George (1987) Long Cycles in World Politics, Seattle: University of Washington Press.

----- (1989) Is America's Decline Inevitable?, The Bridge (Washington DC), 19 (2), Summer, pp.11-18.

----- (1990) "Is world politics evolutionary learning?" International Organization, Vol.44(1), Winter, pp.1-24.

MeEvedy, Colin & Richard Jones (1978) Atlas of World Population History, New York: Penguin.

Perry, Gardner III (1987) "A time series study of developing liberalism in the world", (unpublished m/s).

Reddaway, W.F. et al. (1950) The Cambridge History of Poland, Cambridge: At the University Press.

Rogers, Everett M. (1962) Diffusion of Innovations, New York: Free Press.

Tocqueville, Alexis de (1835) (1969) Democracy in America, (ed. J.P. Mayer), New York: Anchor Books.

Wilson, Woodrow (1918) The State: elements of historical and practical politics, Boston: D.C.Heath.