Many small causes

When large historical events occur, we often want to know the causes that brought them about. And we often look at the world as if these causes too ought to be large, identifiable historical factors or forces. Big outcomes ought to have big, simple causes.

But what if sometimes the historical reality is significantly different from this picture? What if the causes of some "world-historical events" are themselves small, granular, gradual, and cumulative? What if there is no satisfyingly simple and macro answer to the question, why did Rome fall? Or why did the American civil war take the course it did? Or why did North Africa not develop a major Mediterranean economy and trading system? What if, instead, the best we can do in some of these cases is to identify a swarm of independent, small-scale processes and contingencies that eventually produced the outcome?

Take the fall of Rome. I suppose it is possible that the collapse of the empire resulted from a myriad of very different contingencies and organizational features in different parts of the empire: say, logistical difficulties in supplying armies in the German winter, particularly stubborn local resistance in Palestine, administrative decay in Roman Britain, population pressure in Egypt, and a particularly inept series of commanders in Gaul. Too many moving pieces, too much entropy, and some bad luck in personnel decisions, and administrative and military collapse ensues. Alaric sits in Rome.

What an account like this decidedly lacks, is a story about a few key systemic or environmental factors that made collapse "inevitable". Instead, the account is a dense survey of dozens or hundreds of small factors, separated in time and place, whose cumulative but contingent effect was the observed collapse of Rome. No simple necessity here -- "Rome collapsed because of fatal flaw X or environmental pressure Y" -- but instead a careful, granulated assessment of many small and solvable factors.

But here is a different possible historical account of the fall of Rome. An empire depends upon a few key organizational systems: a system of taxation, a system of effective far-flung military power, and a system of local administration in the various parts of the empire. We can take it as a given that the locals will resent imperial taxation, military presence, and governance. So there is a constant pressure against imperial institutions at each locus -- fiscal, military, and administrative. In order to maintain its grip on imperial power, Rome needed to continually support and revitalize its core functions. If taxation capacity slips, the other functions erode as well; but slippage in military capacity in turn undermines the other two functions. And now we're ready for a satisfyingly simple and systemic explanation of the fall of Rome: there was a gradual erosion of administrative competence that led to increasingly devastating failures in the central functions of taxation, military control, and local administration. Eventually this permitted catastrophic military failure in response to a fairly routine challenge. Administrative decline caused the fall of Rome.

I don't know whether either of these stories -- the "many small causes" story or the "systemic administrative failure" story -- is historically credible. But either could be historically accurate. And this is enough to establish the central point: we should not presuppose what the eventual historical explanation will look like.

I suppose there is no reason to expect apriori that large events will conform to either model. It may be that some great events do in fact result from a small number of large causes, while others do not. So the point here is one about the need to expand our historical imaginations, and not to permit our quest for simplicity and generality to obscure the possibility of complexity, granularity, and specificity when it comes to historical causation.

(Christopher Kelly's The Roman Empire: A Very Short Introduction is a very readable treatment of Rome's functioning as an empire. Kelly hands off the ball to Gibbon when it comes to explaining the fall of Rome, however (History of the Decline and Fall of the Roman Empire -- as Kelly says, decidedly not a "short history"). Michael Mann's The Sources of Social Power: Volume 1, A History of Power from the Beginning to AD 1760 gives something of the flavor of my "administrative decline" musing above. Likewise, the style of reasoning about revenues and coercion is very sympathetic to Charles Tilly's arguments about a somewhat later period in Coercion, Capital and European States: AD 990 - 1992.)

Subsistence ethic as a causal factor

In his pathbreaking 1976 book, The Moral Economy of the Peasant: Rebellion and Subsistence in Southeast Asia, James Scott offers an explanation of popular politics based on the idea of a broadly shared "subsistence ethic" among the underclass people of Vietnam and Malaysia. Earlier postings (hidden transcripts, moral economy) have discussed several aspects of Scott's contributions. Here I want to focus on the causal argument that Scott offers, linking the subsistence ethic to the occurrence of rebellion.

Scott's view is that the ensemble of values and meanings current in a society have causal consequences for aggregate facts about the forms of political behavior that arise in that society. Speaking of the peasant rebellions in Southeast Asia of the 1930s Scott writes,

We can learn a great deal from [peasant] rebels who were defeated nearly a half-century ago. If we understand the indignation and rage which prompted them to risk everything, we can grasp what I have chosen to call their moral economy: their notion of economic justice and their working definition of exploitation--their view of which claims on their product were tolerable and which intolerable. Insofar as their moral economy is representative of peasants elsewhere, and I believe I can show that it is, we may move toward a fuller appreciation of the normative roots of peasant politics. If we understand, further, how the central economic and political transformations of the colonial era served to systematically violate the peasantry's vision of social equity, we may realize how a class "of low classness" came to provide . . . the shock troops of rebellion and revolution. (Scott 1976:3-4)

This passage represents a complex explanatory hypothesis about the sources of rebellion. Scott holds, first, that peasant rebels in Indochina in the 1930s shared the main outlines of a sense of justice and exploitation. This is a system of moral values concerning the distribution of material assets between participants (landlord, state, peasant, landless laborer) and the use of power and authority over the peasant. Second, this passage supposes that the values embodied in this sense of justice are motivationally effective: when the landlord or the state enacts policies which seriously offend this sense of justice, the peasant is angered and indignant, and motivated to take action against the offending party. Offense to his sense of justice affects the peasant's actions. Third, Scott asserts that this individual motivational factor aggregates over the peasantry as a whole to a collective disposition toward resistance and rebellion; that is, sufficient numbers of peasants were motivated by this sense of indignation and anger to engage in overt resistance. On this account, then, the subsistence ethic--its right of a subsistence floor and the expectations of reciprocity which it engenders--is a causal antecedent of rebellion. It is a factor whose presence and characteristics may be empirically investigated and which enhances the likelihood of various social events through identifiable mechanisms.

The subsistence ethic may be described quite simply. Scott writes, "we can begin, I believe, with two moral principles that seem firmly embedded in both the social patterns and injunctions of peasant life: the norm of reciprocity and the right to subsistence" (167). Villagers have a moral obligation to participate in traditional practices of reciprocity--labor sharing, contributions to disadvantaged kinsmen or fellow villagers, etc. And village institutions and elites alike have an obligation to respect the right of subsistence of poor villagers.

Claims on peasant incomes . . . were never legitimate when they infringed on what was judged to be the minimal culturally defined subsistence level; and second, the product of the land should be distributed in such a way that all were defined a subsistence niche. (10)

Thus the subsistence ethic functions as a sense of justice--a standard by which peasants evaluate the institutions and persons that constitute their social universe. The subsistence ethic thus constitutes a central component of the normative base which regulates relations among villagers in that it motivates and constrains peasant behavior. And the causal hypothesis is this: Changes in traditional practices and institutions which offend the subsistence ethic will make peasants more likely to resist or rebel. Rebellion is not a simple function of material deprivation, but rather a function of the values and expectations in terms of which the lower class group understands the changes which are imposed upon it.

We can identify a fairly complex chain of causal reasoning in Scott's account. First, the subsistence ethic is a standing condition in peasant society with causal consequences. It is embodied in current moral psychologies of members of the group and in the existing institutions of moral training through which new members are brought to share these values. Through the workings of social psychology this ethic leads individuals to possess certain dispositions to behave. The features and strength of this systems of values are relatively objective facts about a given society. In particular, it is possible to investigate the details of this ethic through a variety of empirical means: interviews with participants, observation of individual behavior, or analysis of the content of the institutions of moral training. Call this ensemble of institutions and current moral psychologies the "embodied social morality" (ESM).

In line with the idea that the subsistence ethic is a standing causal condition, Scott notes that the effectiveness of shared values varies substantially over different types of peasant communities. "The social strength of this ethic . . . varied from village to village and from region to region. It was strongest in areas where traditional village forms were well developed and not shattered by colonialism--Tonkin, Annam, Java, Upper Burma--and weakest in more recently settled pioneer areas like Lower Burma and Cochinchina" (Scott 1976:40). Moreover, these variations led to significant differences in the capacity of affected communities to achieve effective collective resistance. "Communitarian structures not only receive shocks more uniformly but they also have, due to their traditional solidarity, a greater capacity for collective action. . . . Thus, the argument runs, the more communal the village structure, the easier it is for a village to collectively defend its interests" (202).

We may now formulate Scott's causal thesis fairly clearly. The embodied social morality (ESM) is a standing condition within any society. This condition is causally related to collective dispositions to rebellion in such a way as to support the following judgments: (1) If the norms embodied in the ESM were suitably altered, the collective disposition to rebellion would be sharply diminished. (That is, the ESM is a necessary condition for the occurrence of rebellion in a suitable limited range of social situations.) (2) The presence of the ESM in conjunction with (a) unfavorable changes in the economic structure, (b) low level of inhibiting factors, and (c) appropriate stimulating conditions amount to a (virtually) sufficient condition for the occurrence of widespread rebellious behavior. (That is, the ESM is part of a set of jointly sufficient conditions for the occurrence of rebellion.) (3) It is possible to describe the causal mechanisms through which the ESM influences the occurrence of rebellious dispositions. These mechanisms depend upon (a) a model of individual motivation and action through which embodied norms influence individual behavior, and (b) a model of political processes through which individual behavioral dispositions aggregate to collective behavioral dispositions. (That is, the ESM is linked to its supposed causal consequences through appropriate sorts of mechanisms.)

What this account does not highlight -- and what is emphasized by several other theories we've discussed elsewhere (post, post, post, post) -- are the organizational features that underlie successful mobilization. Instead, Scott's account focuses on the motivational features that permit a group to be rallied to the risky business of rebellion.

Causal difference

Source: Federica Russo, Causality and Causal Modelling in the Social Sciences, p. 164

I've recently read a very interesting recent book by Federica Russo, Causality and Causal Modelling in the Social Sciences: Measuring Variations (Methodos Series) on the philosophical issues that arise in causal reasoning about social phenomena. Russo is obviously a talented and dedicated philosopher, and the book is a highly interesting contribution.

Explanation is at the center of scientific research, and explanation almost always involves the discovery of causal relations among factors, conditions, or events. This is true in the social sciences no less than in the natural sciences. But social causes look quite a bit different from causes of natural phenomena. They result from the choices and actions of numerous individuals rather than fixed natural laws, and the causal pathways that link antecedents to consequents are less exact than those linking gas leaks to explosions. Here as elsewhere, the foundational issues are different in the social sciences; so a central challenge for the philosophy of social science is to give a good, compelling account of causal reasoning about social phenomena that does justice to the research problems faced by social scientists. Federica Russo has done so in this book. The book focuses on probabilistic causation and causal modeling, and Russo offers a rigorous and accessible treatment of the full range of current debates. Her central goal is to shed more light on the methods of causal modeling, and she succeeds admirably in this ambition. Causality and Causal Modelling in the Social Sciences makes an important and original contribution.

Her approach to problems in the methodology and philosophy of social science is what she calls the “bottom-up” approach. She works from careful analysis of specific examples of social science reasoning about causation, and works upward to more general analytical findings about causal reasoning as it actually works in the hands of skilled social scientists. She uses five concrete case studies as a vehicle for teasing out the logic of causal inference that is at work: smoking and lung cancer, mother's education and child survival, health and wealth in a population, farmers' migration patterns, and factors causing job satisfaction. She looks at the causal arguments advanced in studies in each of these areas as a way of discovering some of the fundamental logical features of causal inference. The examples are drawn from a wider range of the social and behavioral sciences than is usually the case -- demography, public health, and migration, for example. Her insight is that we can learn a great deal about social science research by looking at good examples of empirical and theoretical reasoning about social changes. I think this is exactly right: it is much better to discover the problems that percolate out of the practice of social inquiry rather than imposing a framework of philosophical expectations onto social science practice.

Russo focuses her attention on the problem of explaining variation: what causes the variation of a certain characteristic over a population of individuals or events. This focus on “variation” rather than “regularity” is very convincing; it provides an appropriate and insightful alternative approach to framing problems for social science research and explanation. She offers a very adept discussion and interpretation of the meaning of causal statements and causal reasoning in the social sciences. The book provides a rigorous contribution to the large literature on the logic of quantitative reasoning about causes of population characteristics. Her case studies are well selected and well done. The book is founded on a deep and rigorous understanding of the most recent philosophical and methodological work on causal modeling.

The author does a very good job of positioning her understanding of the meaning of causal modeling and causal judgments in the social sciences. She honestly addresses the position of “causal realism” and the view that good explanations depend on discovering or hypothesizing causal mechanisms underlying the phenomena. Her chapter on causal mechanisms is a significant contribution to this growing debate within the philosophy of social science; she correctly observes that we need to have greater precision in our discussion of what a causal mechanism is supposed to be.

The book will be of substantial interest to the social scientists, psychologists, demographers, and philosophers who are interested in current debates about the mathematics and philosophy of causal inference. Social scientists and philosophers such as Skyrms, Cartwright, Woodward, Pearl, and Lieberson have developed a very deep set of controversies and debates about the proper interpretation of causal inference. Russo’s book is a substantial contribution to these debates and will engage much the same audiences.

A range of causal questions

In considering important issues in the philosophy of the special sciences, I think it is always helpful to consider a variety of the kinds of intellectual challenges that arise in the area. This gives the philosopher something to work with -- not simply an apriori specification of an issue, but a nuanced set of examples.

So if we are interested in causal reasoning in the social sciences, we ought to pay attention to the kinds of causal questions that social scientists actually want to answer. Let's consider a range of causal questions that have arisen within historical and comparative sociology. In considering these examples, we should reflect on the types of analysis that would provide a satisfactory response to the question, and also the modes of research that would support an empirical response to the question.

• What causes ethnic violence (Horowitz 1985)?
• What caused ethnic violence in Rwanda?
• What caused twentieth-century revolutions (Wolf 1969)?
• What caused the Nicaraguan revolution?
• Why did revolution unfold as it did in the Canton Delta in 1911 (Hsieh 1974)?
• What factors enhance the likelihood of successful democratization (Przeworski 1991; Przeworski et al. 1996)?
• What causes urban residential segregation (Schelling 1978)?
• What causes political corruption (Klitgaard 1988)?
• What factors explain the success or failure of anti-corruption reforms (Klitgaard 1988)?
• What factors explain the East Asian economic miracle (Vogel 1991)?
• Why are there more violent crimes per 1000 in the US than Western Europe?
• Why was the political party of labor more successful in the UK than the US (Przeworski 1985)?
• Why is infant mortality significantly lower in Sri Lanka than Brazil or Egypt (Drèze and Sen 1989, 1995)?
• Why do millenarian cults occur in the post-colonial world (Adas 1979)?
• Why was agricultural technology stagnant in late imperial China (Elvin 1973)?
• Why are rural people more politically conservative than urban people?
• Why do social tastes and styles change as they do (Lieberson 2000)?
• Why did the name “Joshua” lose frequency in the United States in the 1990s (Lieberson 2000)?
• Why did the New England Patriots win the 2003 Super Bowl (Lieberson 1997)?
• Why did the political culture of corporations remain powerful among French workers in the 19th century (Sewell 1980)?
• Why did the heavy wheeled plough diffuse in the geographical pattern that it did in medieval France (Bloch 1966)?
• How did the socialist and republican parties of Spain mobilize the lower working class in support of their programs?
• How did the Solidarity Movement in Poland preserve its organization in face of state repression in the 1980s?
• Was the fact of skewed sex ratios in rural China a necessary condition for the occurrence of banditry and rebellion? Was this fact a contributing condition?
• Why was there no broad-based militant movement of the poor in the United States during the Great Depression?
• Why do restaurants commonly add a gratuity of 18% for parties of 6 or more?
  

We can learn a great deal about causal inquiry by reflecting briefly on a number of these examples. There is a common thread among these examples, in that each topic directs inquiry towards the question, “What are the causal conditions that give rise to a given social or historical outcome?” But there are a number of important differences among these examples as well. Some are about a category of outcome (“twentieth-century revolution” or “ethnic violence”), whereas others are about a historically specific outcome (the Nicaraguan revolution, the Rwandan genocide, the 2003 Super Bowl). Some are about large and publicly salient events, structures, and mentalities (states, revolutions, political cultures); others are about small-scale and unnoticed social characteristics (the frequency of first names). And there are numerous other nuances that emerge from consideration of these examples.

In each case it is a promising research strategy to attempt to discover the underlying social mechanisms that give rise to the outcome -- none of these examples suggests a purely statistical approach to the problem. So inquiry into the "microfoundations" of the causal relations that are uncovered is needed. And second, many of these examples suggest research approaches that make use of the methods of comparative historical sociology and case-study methodology. The techniques of "process-tracing" and small-N comparison of cases should help to arrive at empirically supportable theories of the causal relations that underlie these groups of phenomena.

References
Adas, Michael. 1979. Prophets of rebellion : millenarian protest movements against the European colonial order. Chapel Hill: University of North Carolina Press.
Bloch, Marc Léopold Benjamin. 1966. French rural history; an essay on its basic characteristics. Berkeley,: University of California Press.
Drèze, Jean, and Amartya Kumar Sen. 1989. Hunger and public action. Oxford: Clarendon Press.
———. 1995. India, economic development and social opportunity. Delhi: Oxford University Press.
Elvin, Mark. 1973. The Pattern of the Chinese Past. Stanford: Stanford University Press.
Horowitz, Donald L. 1985. Ethnic Groups in Conflict. Berkeley, California: University of California Press.
Hsieh, Winston. 1974. Peasant Insurrection and the Marketing Hierarchy in the Canton Delta, 1911. In The Chinese City Between Two Worlds, edited by M. Elvin and G. W. Skinner.
Klitgaard, Robert E. 1988. Controlling corruption. Berkeley: University of California Press.
Lieberson, Stanley. 1997. Modeling Social Processes: Some Lessons from Sports. Sociological Forum 12 (1):11-35.
———. 2000. Matter of taste : how names, fashions, and culture change. New Haven, CT: Yale University Press.
Przeworski, Adam. 1985. Capitalism and Social Democracy. Cambridge: Cambridge University Press.
———. 1991. Democracy and the Market: Political and Economic Reforms in Eastern Europe and Latin America, Studies in Rationality and Social Change. Cambridge: Cambridge University Press.
Przeworski, Adam, Michael Alvarez, Jose Antonio Cheibub, and Fernando Limongi. 1996. What makes democracies endure? Journal of Democracy 7 (1).
Schelling, Thomas C. 1978. Micromotives and Macrobehavior. New York: Norton.
Sewell, William Hamilton. 1980. Work and revolution in France : the language of labor from the Old Regime to 1848. Cambridge ; New York: Cambridge University Press.
Vogel, Ezra F. 1991. The Four Little Dragons: The Spread of Industrialization in East Asia. Cambridge, MA: Harvard University Press.
Wolf, Eric R. 1969. Peasant Wars of the Twentieth Century. New York: Harper & Row.

Causing public opinion

It is interesting to consider what sorts of things cause shifts in public opinion about specific issues. This week's national election is one important example. But what about more focused issues -- for example, the many ballot initiatives that were considered in many states? To what extent can we discover whether there is a measurable effect on public opinion by the organized efforts of advocacy groups through advertising and other strategies for reaching the minds of voters?

In these cases we might imagine that voters have a prior set of attitudes towards the issue -- perhaps including a large number of "don't know/don't care" people. Then a set of advocates form to lobby the public pro and con. They mount campaigns to influence voters' opinions towards the option they prefer. And on the day of the election voters will indicate their approval -- often in ratios quite different from those that were measured in pre-campaign surveys. So something happened to change the composition of public opinion on the issue. The question here is whether it is possible to estimate the effects of various possible influencers.

This seems like potentially a very simple area of causal reasoning about social processes. The outcome variable is fairly observable through polling and the final election, and the interventions are also usually observable as well, both in timing and magnitude. So the world may present us with a series of interventions and outcomes that support fairly strong causal conclusions -- for example, "each time ad campaign X hits the airwaves in a given market, there is an observed uptick in support for the proposition." It is unlikely that the correlation occurred as a result of random variations in both terms; we have a theory of how advertising influences voters; and we conclude that "ad campaign X was a causal factor in shaping voter opinion in this time period." (It is even possible that X played a role in both segments of opinion, resulting in an up-tick in both yes and no responses. Then we might also judge that X was effective at polarizing voters -- not the effect the strategist would have aimed at.)

This is an example of singular causal reasoning, in that it has to do with one population, one issue, and a specific series of interventions. What would be needed in order to arrive at a conclusion with generic scope -- for example, "advertising along the lines of X is generally effective in increasing support for its issue"? The most straightforward argument to the generic conclusion would be a study of an extended set of cases with a variety of strategies in play. If we discover something like this -- "In 80% of cases where X is included in the mix it is observed to have a positive effect on opinion" -- then we would have inductive reason for accepting the generic causal claim as well. This is basic experimental reasoning.

Take a hypothetical issue -- a referendum on a proposal for changing the system the state uses for assessing business taxes. Suppose that a polling firm has done weekly polling on the question and has recorded "yes/no/no opinion" since October 2007. Suppose that two organizations emerged in December to advocate for and against the proposal; that each raised about $5 million; and that each included an advertising campaign in its strategy. Suppose further that the "no" campaign also included a well-organized effort at the parish level to persuade church members to vote against the measure on religious grounds and the "yes" campaign included a grassroots effort to get university students and staff to be supportive of the measure on pro-science and pro-economy grounds. And suppose each organization mounted a "new media" campaign using email lists and web comminication to make its case. Finally, suppose we have good timeline data about the occurrence and volume of media spots throughout the period of June through November.

This scenario involves three types of causes, a timeline representing the application of the interventions, and a timeline representing the effects. From this body of data can we arrive at estimates of the relative efficacy of the three treatments? And does this set if conclusions provide credible guidance for other campaigns over other issues in other places?

There is also the question of the efficacy of the implementation of the strategies. Take the ad campaigns. Whether a specific campaign succeeds in changing viewers' opinions depends on the content, message, and production quality. Does the message resonate with a target segment of voters? Does the production design stimulate emotions that will lead to the desired vote? So evaluating efficacy needs to be done across instances of media as well as across varieties of media. (This is the function of focus groups and snap polls -- to evaluate the effects of specific messages and production choices on real voters.)

(Here is a link to some information about the process leading up to a positive vote on the Michigan Stem Cell initiative this month. A good general introduction to the social psychological theories about the formation of attitudes and opinions is Stuart Oskamp and P. Wesley Schultz, Attitudes and Opinions.)

Policy, treatment, and mechanism

Policies are selected in order to bring about some desired social outcome or to prevent an undesired one. Medical treatments are applied in order to cure a disease or to ameliorate its effects. In each case an intervention is performed in the belief that this intervention will causally interact with a larger system in such a way as to bring about the desired state. On the basis of a body of beliefs and theories, we judge that T in circumstances C will bring about O with some degree of likelihood. If we did not have such a belief, then there would be no rational basis for choosing to apply the treatment. "Try something, try anything" isn't exactly a rational basis for policy choice.

In other words, policies and treatments depend on the availability of bodies of knowledge about the causal structure of the domain we're interested in -- what sorts of factors cause or inhibit what sorts of outcomes. This means we need to have some knowledge of the mechanisms that are at work in this domain. And it also means that we need to have some degree of ability to predict some future states -- "If you give the patient an aspirin her fever will come down" or "If we inject $700 billion into the financial system the stock market will recover."

Predictions of this sort could be grounded in two different sorts of reasoning. They might be purely inductive: "Clinical studies demonstrate that administration of an aspirin has a 90% probability of reducing fever." Or they could be based on hypotheses about the mechanisms that are operative: "Fever is caused by C; aspirin reduces C in the bloodstream; therefore we should expect that aspirin reduces fever by reducing C." And ideally we would hope that both forms of reasoning are available -- causal expectations are born out by clinical evidence.

Implicitly this story assumes that the relevant causal systems are pretty simple -- that there are only a few causal pathways and that it is possible to isolate them through experimental studies. We can then insert our proposed interventions into the causal diagram and have reasonable confidence that we can anticipate their effects. The logic of clinical trials as a way of establishing efficacy depends on this assumption of causal simplicity and isolation.

But what if the domain we're concerned with isn't like that? Suppose instead that there are many causal factors and a high degree of causal interdependence among the factors. And suppose that we have only limited knowledge of the strength and form of these interdependencies. Is it possible to make rationally justified interventions within such a system?

This description comes pretty close to what are referred to as complex systems. And the most basic finding in the study of complex systems is the extreme difficulty of anticipating future system states. Small interventions or variations in boundary conditions produce massive variations in later system states. But this is bad news for policy makers who are hoping to "steer" a complex system towards a more desirable state. There are good analytical reasons for thinking that they will not be able to anticipate the nature or magnitude or even direction of the effects of the intervention.

The study of complex systems is a collection of areas of research in mathematics, economics, and biology that attempt to arrive at better ways of modeling and projecting the behavior of systems with these complex causal interdependencies. This is an exciting field of research at places like the Santa Fe Institute and the University of Michigan. One important tool that had been extensively developed is the theory of agent-based modeling -- essentially, the effort to derive system properties as the aggregate result of the activities of independent agents at the micro-level. And a fairly durable result has emerged: run a model of a complex system a thousand times and you will get a wide distribution of outcomes. This means that we need to think of complex systems as being highly contingent and path-dependent in their behavior. The effect of an intervention may be a wide distribution of future states.

So far the argument is located at a pretty high level of abstraction. Simple causal systems admit of intelligent policy intervention, whereas complex, chaotic systems may not. But the important question is more concrete: which kind of system are we facing when we consider social policy or disease? Are social systems and diseases examples of complex systems? Can social systems be sufficiently disaggregated into fairly durable subsystems that admit of discrete causal analysis and intelligent intervention? What about diseases such as solid tumors? Can we have confidence in interventions such as chemotherapy? And, in both realms, can the findings of complexity theory be helpful by providing mathematical means for working out the system effects of various possible interventions?

What causes college success?

This sounds like a simple question. It sounds as if it is asking for us to discover a set of factors that influence the level of performance of individuals within a population when they get to colleges and universities. And we might speculate that there is a small group of potentially relevant factors: antecedent cognitive ability, attitudes, and values; location within a set of social relations that enhance or impede successful educational performance; quality of educational resources provided in K-12. We might reason that a given individual's performance is affected by his/her ability and motivation; enhancing or inhibiting circumstances; quality of educational "treatment"; and chance events or circumstances (a lucky break, an inspiring grandfather). And by examining antecedent conditions and outcomes across a large population of people, we might expect to be able to assess the degree to which various hypothesized factors in fact lead to differences in the performance of sub-populations defined by these factors. This analysis should shed light on the question, "What factors cause differences in university success?".

Sorting this out sounds like a straightforward empirical question. Consider this hypothetical study. First, identify a cohort of high school seniors -- let's say, all the seniors in 2000 in metropolitan Boston. Suppose this is 5,000 people. (1) Measure a set of features of their situation during high school: high school performance, family situation, features of the school attended, socioeconomic status, family status, racial-ethnic status. (2) Measure a set of psychological characteristics for each individual: motivation, determination, aptitude for mathematics, ... And, (3), measure college success five years following high school graduation (GPA, credit hours completed, degree attained).

Let's say that each individual is coded for ABILITY (1, 2, 3); MOTIVATION (1, 2, 3); SOCIOECONOMIC STATUS (1, 2, 3); RACE (A, B, C); FAMILY STATUS (A, B); HIGH SCHOOL QUALITY (1, 2, 3); HIGH SCHOOL PERFORMANCE (1, 2, 3); and a factor representing one particular educational or curricular theory -- let's say, PEER COUNSELING (T,F). And let's say that outcomes are coded as DEGREE (NONE, ASSOCIATES, BACHELORS, MASTERS, PROFESSIONAL) and GPA (1, 2, 3).

Now follow these individuals for 10 years: What further education do they pursue? Do they complete post-secondary education? What is their performance in post-secondary education? What occupations and jobs do they get? What income do they achieve by age 30? How much unemployment have they experienced?

Finally, we will do some basic statistics on this data set: compute the incomes and schooling for various sub-categories; test for correlations between outcomes and antecedent conditions; etc. Are there differences in outcomes when we cross-tabulate by ABILITY or MOTIVATION? What about if we cross-tabulate by RACE or SES? This analysis may produce statements like these hypothetical findings:

• People who completed high school with high performance were 2.5 times as likely to complete a college degree as those with a low performance.
• People whose family income was in the top quintile were 5 times as likely to complete a college degree as those from families in the bottom quintile.
• The college completion rate for white students, Hispanic students, and African-American students were X, Y, and Z respectively.
• High school graduates from high schools with peer counseling programs were X percent more likely to complete a bachelor's degree.
  
• People living in single-parent households during high school had completion rates of X compared to Y for dual-parent households.

A study along these lines provides a first indication of how some of these social characteristics may be related to performance in college. If a factor is not causally related to the outcome, then the population possessing this factor should have the same performances as the population lacking this factor (the null hypothesis). So if we find that differences in family structure or performance in high school are associated with differences in college performance, then we can infer that these factors play some causal or structural role in the outcome.

However, these findings do not establish specific causal linkages among the factors. Take the hypothetical finding about family income: is this statistical discovery the result of this mechanism (greater family income provides more support for tutoring and academic support) or this mechanism (greater family income is associated with familial values that put strong emphasis on successful completion of university degree) or this mechanism (greater family income confers social advantages that make completion easier for affluent students)? In other words, the statistical discovery does not determine the nature of the causal relation between the antecedent condition and the outcome; it simply points the researcher towards investigating the concrete social mechanisms that might be at work here.

The example demonstrates an important lesson about social inquiry. Statistical study of a population can in fact point us towards some preliminary hypotheses about social causation. But these statistical discoveries are only the first step. In order to confidently assert causal relationships between things like income and race, to educational outcomes, we need to arrive at a nuanced analysis of the social relations and institutions through which these gross factors play into individual outcomes. We need to have an account of the mechanisms and processes through which the effects of concrete social settings characterized by differences in family structure, SES, race, or schools play out in the social psychology and educational opportunities that determine the ultimate outcomes of the young people who pass through them.

(A similar line of thought can be found in this posting on the problem of sorting out the data establishing correlations between race and asthma.)

Quasi-experimental data?

Stan Lieberson is one of a group of sociologists for whom I have great respect when it comes to intelligent thinking about social science methodology. His 1985 book, Making It Count: The Improvement of Social Research and Theory, is a good example of some of this thinking about the foundations of social science knowledge, and I also admire A Matter of Taste: How Names, Fashions, and Culture Change in the way it offers a genuinely novel topic and method of approach.

Lieberson urges us to consider "a different way of thinking about the rigorous study of society implied by the phrase 'science of society'" instead of simply assuming that social science should resemble natural science (3-4). His particular object of criticism in this book is the tendency of quantitative social scientists to use the logic of experiments to characterize the data they study.

An experiment is an attempt to measure the causal effects of one factor X on another factor Z by isolating a domain of phenomena -- holding constant all other causal factors -- and systematically varying one causal factor to observe the effect this factor has on an outcome of interest. The basic assumption is that an outcome is the joint effect of a set of (as yet unknown) causal conditions:

C1 & C2 & ... & Cn cause Z,

where we do not yet know the contents of the list Ci. We consider the hypothesis that Cm is one of the causes of Z. We design an experimental environment in which we are able to hold constant all the potentially relevant causal conditions we can think of (thereby holding fixed Ci), and we systematically vary the presence or absence of Cm and observe the state of the outcome Z. If Z varies appropriately with the presence or absence of Cm, we tentatively conclude that Cm is one of the causes of Z.

In cases where individual differences among samples or subjects may affect the outcome, or where the causal processes in question are probabilistic rather than deterministic, experimentation requires treating populations rather than individuals and assuring randomization of subjects across "treatment" and "no-treatment" groups. This involves selecting a number of subjects, randomly assigning them to controlled conditions in which all other potential causal factors are held constant, exposing one set of subjects to the treatment X while withholding the treatment from the other group, and measuring the outcome variable in the two groups. If there is a significant difference in the mean value of the outcome variable between the treatment group and the control group, then we can tentatively conclude that X causes Z and perhaps estimate the magnitude of the effect. Take tomato yields per square meter (Z) as affected by fertilizer X: plants in the control group are subjected to a standard set of growing conditions, while the treatment group receives these conditions plus the measured dose of X. We then measure the quantity produced by the two plots and estimate the effect of X. The key ideas here are causal powers, random assignment, control, and single-factor treatment.

However, Lieberson insists that most social data are not collected under experimental conditions. It is normally not possible to randomly assign individuals to groups and then observe the effects of interventions. Likewise, it is not possible to systematically control the factors that are present or absent for different groups of subjects. If we want to know whether "presence of hate speech on radio broadcasts" causes "situations of ethnic conflict" to progress to "situations of ethnic violence" -- we don't have the option of identifying a treatment group and a control group of current situations of ethnic conflict, and then examine whether the treatment with "hate speech on radio broadcasts" increases the incidence of ethnic violence in the treatment group relative to the control group. And it is fallacious to reason about non-experimental data using the assumptions developed for analysis of experiments. This fallacy involves making "assumptions that appear to be matters of convenience but in reality generate analyses that are completely off the mark" (6).

Suppose we want to investigate whether being a student athlete affects academic performance in college. In order to treat this topic experimentally we would need to select a random group of newly admitted students; randomly assign one group of individuals to athletic programs and the other group to a non-athletic regime; and measure the academic performance of each individual after a period of time. Let's say that GPA is the performance measure and that we find that the athlete group has a mean GPA of 3.1 while the non-athlete group has an average of 2.8. This would be an experimental confirmation of the hypothesis that "participation in athletics improves academic performance."

However, this thought experiment demonstrates the common problem about social data: it is not possible to perform this experiment. Rather, students decide for themselves whether they want to compete in athletics, and their individual characteristics will determine whether they will succeed. Instead, we have to work with the social realities that exist; and this means identifying a group of students who have chosen to participate in athletics; comparing them with a "comparable" group of students who have chosen not to participate in athletics; and measuring the academic performance of the two groups. But here we have to confront two crucial problems: selectivity and the logic of "controlling" for extraneous factors.

Selectivity comes in when we consider that the same factors that lead a college student to participate in athletics may also influence his/her academic performance; so measuring the difference between the two groups may only measure the effects of this selective difference between membership in the groups -- not the effect of the experience of participating in athletics on academic performance. In order to correct for selectivity, the researcher may attempt to control for potentially influential differences between the two groups; so he/she may attempt to control for family factors, socio-economic status, performance in secondary school, and a set of psycho-social variables. "Controlling" in this context means selecting sub-groups within the two populations that are statistically similar with respect to the variables to be controlled for. Group A and Group B have approximately the same distribution of family characteristics, parental income, and high school GPA; the individuals in the two groups are "substantially similar". We have "controlled" for these potentially relevant causal factors -- so any observed differences between academic performance across the two groups can be attributed to the treatment, "participation in athletics."

But Lieberson makes a critical point about this approach: there is commonly unmeasured selectivity within the control variables themselves -- crudely, students with the same family characteristics, parental income, and high school GPA who have selected athletics may nonetheless be different from those who have not selected athletics, in ways that influence academic performance. As Lieberson puts the point, "quasi-experimental research almost inevitably runs into a profound selectivity issue" (41).

There is lots more careful, rigorous analysis of social-science reasoning in the book. Lieberson crosses over between statistical methodology and philosophy of social science in a very useful way, and what is most fundamental is his insistence that we need to substantially rethink the assumptions we make in assigning causal influence on the basis of social variation.

Is sociology analogous to epidemiology?

Quantitative sociology attempts, among other things, to establish causal connections between large social factors (race, socio-economic status, residential status) and social outcomes of interest (rates of delinquency). Is this type of inquiry analogous in any way to the use of large disease databases to attempt to identify risk factors? In other words, is there a useful analogy between sociology and epidemiology?

Suppose that the divorce rate for all American men is 30%. Suppose the rate for New York City males with income greater than $200K is 60%. We might want to draw the inference that something about being a high-income male resident of New York causes a higher risk of divorce for these persons. And we might want to justify this inference by noticing that it is similar to a parallel statistical finding relating smoking to lung cancer. So sociology is similar to epidemiology. Certain factors can be demonstrated to cause an elevated risk of a certain kind of outcome. There are "risk factors" for social outcomes such as divorce, delinquency, or drug use.

Is this a valid analogy? I think it is not. Epidemiological reasoning depends upon one additional step -- a background set of assumptions about the ontology and etiology of disease. A given disease is a specific physiological condition within a complex system of cells and biochemical processes. We may assume that each of these physiological abnormalities is caused by some specific combination of external and internal factors through specific causal mechanisms. So the causal pathways of normal functioning are discrete and well-defined, and so are the mechanisms that cause disruption of these normal causal pathways. Within the framework of these guiding assumptions, the task of the statistics of epidemiology is to help sort out which factors are causally associated with the disease. The key, though, is that we can be confident that there is a small number of discrete causal mechanisms that link the factor to the disease.

The case is quite different in the social world. Social processes are not similar to physiological processes, and social outcomes are not similar to diseases. In each case the failure of parallel derives from the fact that there are not unique and physiologically specific causal systems at work. Cellular reproduction has a specific biochemistry. Cancerous reproduction is a specific deviation from these cellular processes. And specific physical circumstances cause these deviations.

Now think about the social world. A process like "urbanization" is not a homogeneous social process. Rather, it is a heterogeneous mix of social developments and events; and these components are different in different times and places. And outcomes that might be considered the social equivalent of disease -- a rising murder rate, for example -- is a also composite of many distinct social happenings and processes. So social systems and outcomes lack the simple, discrete causal uniformity that is a crucial part of epidemiological reasoning.

This is not to say that there are not underlying causal mechanisms whose workings bring about a sharp increase in, say, the population's murder rate. Rather, it is to say that there are numerous, heterogeneous and cross-cutting such mechanisms. So the resultant social outcome is simply the contingent residue of the multiple middle-level processes that were in play in the relevant time period. And the discovery that "X, Y, Z factors are correlated with a rise in the incidence of O" isn't causally irrelevant. But the effects of these factors must be understood as working through their influence on the many mid-level causal mechanisms.

Paired comparisons

Sidney Tarrow is a gifted and prolific student of comparative politics. (Listen to my interview with Professor Tarrow.) He has spent much of his career trying to understand social movements, contentious politics, and the causes of differences in political behavior across national settings. And one of his special contributions is his ability to think clearly about the methods that social scientists use.

Tarrow attaches a lot of weight to the idea of "paired comparisons" as a method of research and discovery: Locate a few cases that are broadly similar in many respects but different in a way that is important, interesting, or surprising. Then examine the cases in greater detail to attempt to discover what explains the difference between the two cases. (One of his early books that employs this method is From center to periphery: Alternative models of national-local policy impact and an application to France and Italy.)

Nothing special turns on "pairs" here; what Tarrow is describing is really the logic of small-N comparative research. The point about the broad similarity that is the basis for choosing the cases follows from the logic of causation: if we presuppose that the outcome P is caused by some set of antecedent social and political conditions and we know that C1 and C2 have different outcomes -- then the more factors we can "control" for by finding cases in which these factors are constant, the better. This is so, because it demonstrates that none of the constant factors in the two cases are the cause of variation in outcome. And this limits our investigation of possible causes to the factors in which the cases differ.

If this sounds like Mill's methods of similarity and difference, that's appropriate -- the logic is the same, so far as I can see. Here is Mill's method of difference:

A B C D -E => P
A B -C D -E => -P

And in this case -- making the heroic assumption that A,B,C,D,E exhaust all possible causes of P, and that the cause of P is deterministic rather than probabilistic -- then we can infer that the presence of C causes P.

This reasoning doesn't take us to a valid conclusion to the effect that C is the only factor that is causally relevant to the occurrence of P; it is possible, for example, that there is a third case along these lines:

-A B -C D -E => -P

This would demonstrate that A is a necessary condition for the occurrence of P; withhold A and P disappears. And each of the other factors might also play a role as a necessary condition. So it would be necessary to observe as many as 32 cases (2^5) in order to sort out the status of A through E as either necessary or sufficient conditions for the occurrence of P. (The logic of this kind of causal reasoning is explored more fully in my essay, "An Experiment in Causal Reasoning," which is also published in Microfoundations, Methods, and Causation.)

But I don't think that Tarrow is intending to advance the method of paired comparison as a formal method of causal inference, along the lines of inductive or deductive logic. Instead, I think he is making the plausible point that this method should be understood as a part of an intelligent research strategy. Social processes are complex. We are interested in explaining variation across cases. And we probably have the best likelihood of discovering important causal relationships if we can reduce the number of moving parts (the other kinds of variation that occur across the cases).

Tarrow gives an example of the application of the method of paired comparisions in the context of his early study of the political fortunes of the Italian Communist Party (PCI) in the south of Italy. In this case the paired comparison involves northern Italy and southern Italy. Both are subject to the same national political structures; both populations speak Italian; both populations have an Italian national identity. However, the PCI was fairly successful in mobilizing support and winning elections based on a militant political program in the north, and was persistently unsuccessful in doing these things in the south. What explains the difference?

As Tarrow explains his reasoning, his expectation in conducting the research was a "structural" one. He expected that there would be large structural factors in post-war Italy -- features of economic and political institutions -- that would explain the difference in outcome for PCI political activism. And there were indeed large structural differences in social organization in the two regions. Northern Italy possessed an economy in which industrial labor played a key role and constituted a substantial part of the population. Southern Italy was agrarian and backward, with a large percentage of exploited peasants and only a small percentage of industrial workers.

But, very significantly, Tarrow now believes that these "structural" expectations are probably too "macro" to serve as the basis of social explanation. Instead, he favors the importance of looking at the dynamics of social processes and the specific causal mechanisms that can be discovered in particular social-historical settings. This means looking for causal factors that work at a more strategic and meso level. In terms of the southern Italian PCI outcome that he was interested in explaining thirty years ago -- he now believes that the causal mechanism of "brokerage" would have shed much more light on the political outcomes that were of interest in Italy. (This is the heart of the approach that he takes, along with Doug McAdam and Chuck Tilly, in Dynamics of Contention.)

This finding doesn't invalidate the heuristic of paired comparisons. But it probably does invalidate the expectation that we might discover large "structure-structure" patterns of causation through such comparisons. Instead, what the method facilitates is a more focused research effort on the part of the comparativist, in the context of which he/she can search out the lower-level causal mechanisms and processes that are at work in the various settings under study.