Application of Boolean Methods of Qualitative Comparison Abridged from Ragin Charles, The Comparative Method

Download 42.87 Kb.
Date conversion04.05.2016
Size42.87 Kb.
Application of Boolean Methods of Qualitative Comparison

Abridged from Ragin Charles, The Comparative Method, University of California Press, 1987.

The application presented here is a re-analysis of data used by Rokkan (1970) in his work on nation building in Western Europe. Rokkan used a "configurational" approach that bears many similarities to the Boolean approach presented in this work. His main substantive interest was the growth of mass democracy and the emergence of different cleavage structures in Western European polities. One outcome that interested him was the division of some working-class movements in these countries following the Russian Revolution into internationally oriented wings and some into nationally oriented wings. He considered the distribution of this outcome important because of its implication for the future of working-class mobilization (and cleavage structures in general) in Western Europe.
Many of the methodological sentiments expressed in this course echo those voiced by Stein Rokkan in his pioneering work on nation building published in the late 1960s and early 1970s. Rokkan was disturbed by the gulf between case-oriented and variable-oriented study and proposed an explicitly configurational approach to comparative social research as a way to bridge the two strategies. The research strategy he outlined resembles the Boolean approach presented in this book in its emphasis on combinations of characteristics and holistic comparison of cases.
In a typical application of his configurational approach, Rokkan would establish three or four theoretically important dichotomies and then elaborate their different logically possible combinations. Countries manifesting each combination of values would then be selected, compared, and interpreted. These results, in turn, would be used as a basis for evaluating the heuristic value of the conceptual framework represented in the dichotomies. If the empirical examples of the different combinations of characteristics differed in predicted ways from each other, this was taken as evidence in favor of the value of the scheme as a guide to historical interpretation.
One of the issues that especially interested Rokkan was the timing and speed of the extension of the franchise (franchise = the right to vote) in Western European countries and, by implication, the amount of conflict associated with the growth of mass democracy in each country. Three historical conditions defining different starting points in this process, he argued, shaped the progress of democratization:
First, "medieval consolidation" whether the country was a separate dynasty or a collection of cities and provinces within successive continental empires;
Second, "continuity of representative organs" whether or not the country experienced extensive periods of absolutist rule;
Third, "status in the international system" whether a country was, or was part of, a major power or a lesser power.
After examining the extension of the franchise in cases representative of each combination of values (there were only a few combinations lacking empirical instances), Rokkan concluded that the character of franchise extension was indeed shaped by different combinations of these three historical conditions.
Often, Rokkan's configurational approach had a somewhat nebulous quality to it. In the example cited above, the dependent variable (Y) was the character of the growth of mass democracy. Thus, the analysis examined different historical conditions shaping the nature of this growth, not any particular feature of it. This aspect of Rokkan's work tilts it in a holistic, case-oriented direction - despite the generalizing, variable-oriented character that follows from applying the same framework to a range of cases.
Occasionally, however, Rokkan did address specific historical outcomes. One feature of the history of Western European polities that interested him, for example, was the variation among them in the impact of the Russian Revolution on working-class organizations. In some countries it had little impact, but in others it created deep and lasting divisions. A cursory examination of the cross-national distribution of these divisions does not yield simple conclusions. For example, Sweden and Norway are neighboring countries and share many features. Yet the success of the Russian Revolution, according to Rokkan, created only minor divisions in Swedish working-class organizations but major divisions in Norwegian organizations. True to form, Rokkan addressed this variation configurationally. In essence, he argued that the origins and nature of a polity's existing cleavage structure shaped the reaction of a country's working-class movement to the Russian Revolution.
It would be difficult, of course, to reproduce his entire argument on cleavage structures in this brief treatment. His main concern was the interests and alliances of the state-builders and how these factors shaped the nature of the opposition to the state-builders. Of necessity, these interests and alliances were historically grounded. Rokkan argued that the important historical factors shaping cleavage structures in Western European polities and their reactions to the Russian Revolution were the outcome of the Reformation, the outcome of the "Democratic Revolution", the outcome of the Industrial Revolution, and the timing of state formation. The important dichotomies related to these four factors were:
1. Whether the state established a national church or remained allied with the Roman Catholic Church. Rokkan labels this outcome "C" for national church (x1)
2. Whether or not the state allowed Roman Catholic participation in nation-building institutions, especially mass education (x2). In countries with national churches, this indicates deep religious division. In countries that remained allied with the Roman Catholic Church, this represents a failure to establish a more secular state. Obviously, this dichotomy is relevant only to countries with large numbers of Roman Catholics. Rokkan labels this outcome "R" for Roman Catholic.
3. Whether the state maintained an alliance with landed interests or favored commercial and industrial interests over landed interests from the outset (x3). Rokkan labels this outcome "L" for landed interests.
4. Whether a state formed early (such as Spain) or late (such as Belgium) (x4). Rokkan labels this outcome "E" for early.
These four dichotomies yield sixteen different combinations of conditions. Rokkan identified empirical instances of ten of these combinations (See table below.)
The outcome variable in the table below is labeled "S" and indicates working-class parties that were split in their reaction to the Russian Revolution.

After examining the different combinations of conditions and their associated outcomes, Rokkan (1970:132-138) concludes that in Protestant countries (that is, those with national churches) the working-class movement tended to be much more divided if the nation-building process was more recent and, by implication, national identity less settled. In Catholic countries, by contrast, the deeper and more persistent the church-state conflict, the greater the division in the working-class movement. In general, it appears from these two combinations that the less settled polities (Protestant ones, which are newer; Catholic ones because of continuing religion conflict) were the ones that experienced divided working-class movement.

Rokkan's Data on Divided Working-Class Movements












Great Britain






No Instance







C = National church (vs. state allied to Roman Catholic church)

R = Significant Roman Catholic population and Roman Catholic participation in mass education

L = State protection of landed interests

E = Early state

S = Major split in working-class movement provoked by Russian Revolution (outcome variable)

* Question marks indicate that no clear prediction is made

It is easy to express Rokkan's conclusion in Boolean terms (with uppercase letters indicating presence and lowercase letters indicating absence), and it is roughly confirmed through simple inspection of the empirical data presented in truth table form in the table above.

S = Ce + cr
The equation states simply that the Russian Revolution divided working-class movements (1) in countries with national churches that had experienced nation building more recently (Norway, Finland, Iceland, Germany) and (2) in countries without national churches (that is. Catholic countries) that had denied the Roman Catholic church a major role in mass education (Spain, France, Italy).
Rokkan's results are duplicated only if the combinations of conditions without empirical instances (the last six rows in the table) are allowed to take on any output value. In this type of analysis, the algorithm may assign these rows l's or 0's, whichever assignment produces the most logically minimal solution possible. This is equivalent to incorporating simplifying assumptions that, in effect, make allowances for the limited diversity of social phenomena (in this case, the limited diversity of Western European countries).
Boolean analysis of Rokkan's data without these simplifying assumptions does not reproduce his results. The most conservative way to approach the data in the truth table is to assume that the six combinations of characteristics for which there are no empirical instances would not have divided working-class movements. This strategy is conservative only in the sense that it treats the division of the working class as an unusual phenomenon and, by implication, considers no division following the Russian Revolution (a likely consequence of sheer inertia) the normal state of affairs. This assumption is operationalized simply by coding the output for these six combinations of values to zero in the table.
Applying the Boolean minimization algorithms to the resulting truth table yields the following reduced, expression:
S = rle + crE + CRLe
This equation is considerably more complex 'than the one allowing simplifying assumptions (that is, Rokkan's). It describes three different (mutually exclusive) combinations of conditions leading to divided working-class movements: (1) low Roman Catholic involvement in mass education in a more recently formed state that favored urban interests from the outset: Italy, Norway, Finland, and Iceland; (2) low Roman Catholic involvement in mass education in a Catholic country with a long history of state building: Spain and France; and (3) Roman Catholic involvement in mass education in a Protestant country with a recent history of state building allied with landed interests: Germany.
The two conditions identified by Rokkan (Ce and cr) are clearly visible in the last two terms of the second equation. Thus, the second and third terms in this equation could be considered elaborations of his basic argument which emphasized recency in Protestant countries and religiously based conflict in Catholic countries. Note, however, that the last term (the one relevant to his Ce combination) also includes religious conflict - Catholic involvement in mass education in a Protestant country. Thus, these elaborations of Rokkan's simpler terms give greater weight to a history of religious conflict. In many respects, therefore, both of these terms describe national situations where the pressure or weight of historically rooted conditions on political institutions and arrangements was great. (In many respects, the weight of history was comparably great in Russia.)
Considering these two terms alone, there is some resonance of the results with arguments made by Mann (1973) and echoed by Giddens (1973). Mann and Giddens present elaborate historical arguments concerning conditions that prompt the development of revolutionary working-class consciousness. They both argue that where the confrontation between a feudal past and modern institutions was most sudden and acute, revolutionary consciousness was most likely. To the extent that a divided working-class movement signals a greater reservoir of potential revolutionary consciousness, this argument is loosely supported by the last two terms in the equation.
The first term in the equation, however, is not consistent with Rokkan's argument or with the argument concerning the weight of historical cleavages developed above. The image conveyed by this combination is of a highly secular state (whether it is Protestant or Catholic is irrelevant) that is relatively free from historical constraints: it is not allied with landed interests, nor is it encumbered by historically rooted political institutions. This combination of conditions casts a very different light on the question of reactions to the Russian Revolution. It suggests that the Russian Revolution had a strong impact on polities (and working-class movements) that were less constrained by historical cleavages and more open to change. In short, the inertia of the past was easier to overcome in these cases.
Together, the three terms in the equation suggest that divided working-class movements were found in countries where the burden of historically rooted conditions on the polity was either relatively light or very heavy. This conclusion is qualitatively different from Rokkan's, which emphasized the degree to which different polities were "settled." Of course, this generalization is limited to Western Europe after the Russian Revolution. It would be hazardous to extend this statement beyond this region and period.

There is still another way to evaluate Rokkan's analysis. I noted above that if the six combinations lacking empirical instances are allowed to take on any output value, then it is possible to reproduce Rokkan's conclusion (S = cr+ Ce) with Boolean techniques. However, this simpler solution requires simplifying assumptions. The important question to answer from this perspective is "what was Rokkan required to assume in order to produce this tidy solution?" This can be ascertained by contrasting the first solution, which incorporates simplifying assumptions, with the second, which does not.

An analysis of these differences shows that Rokkan assumed - implicitly - that countries with the following combinations of conditions, if they had existed, would have experienced divided working-class movements following the Russian Revolution: CRle, rLe. The first term describes a more recently formed Protestant nation-state with heavy Roman Catholic involvement in mass education and a bias toward urban interests. The second describes a recently formed nation-state with a bias toward landed interests and with low Catholic involvement in mass education. Of course, there are no clear instances of these two combinations within Western Europe, and Rokkan did not intend his argument to be applied outside this region.

However, there are countries that roughly approximate these combinations outside of Western Europe, and these cases could be examined to see if they are consistent with Rokkan's expectations. This examination would provide an avenue for establishing a crude check on Rokkan's simplifying assumptions.

The important point is not that these cases were not checked but that simplifying assumptions were implicitly incorporated. Boolean techniques provide a direct avenue for uncovering simplifying assumptions, which makes it possible to bring them forward for examination.
The intent of this application has not been to criticize Rokkan but simply to show how Boolean methods elaborate his configurational approach. Rokkan indicated that his conclusions were tentative. The ones offered here based on his classifications are even more tentative than Rokkan's. Rokkan's primary goal was to establish a foundation for examining the development and structure of cleavage systems in Western Europe. If anything, the re-analysis offered here simply confirms that the scheme he developed is useful, perhaps in ways he did not intend. Nevertheless, the goal of the re-analysis is compatible with Rokkan's to provide a foundation for understanding historical patterns and political developments in Western European polities, not to test theory per se.

1. What is the difference between Ragin's and Rokkan's theses about the sources of divided working class movement?.

2. To what extent they express the Israeli case of divided working class movement?

2. What are the advantages of Ragin's formulation?.

3. What are the sources of Rokkan oversimplification assumption?.

Giddens, Anthony, The Class Structure of the Advanced Societies, London, Hutchinson, 1973.
Mann, Michael, Consciousness and Action Among the Western Working Class, London, Humanities, 1973.
Rokkan, Stein, Citizens, Elections, Parties. New York: McKay, 1970.

The database is protected by copyright © 2016
send message

    Main page