Transcript PowerPoint Presentation - Topics in the Philosophy of

Causal realism
Daniel Little
University of
Michigan-Dearborn
August 6, 2007
www-personal.umd.umich.edu/~delittle/
1
I. Causal explanation
• Explanations in the social sciences are
almost invariably causal explanations: to
explain the outcome or the regularity, we
undertake to discover the causes /
conditions / circumstances that combine to
bring the outcome about.
• What is a causal explanation?
2
Manifesto
• There are causal relations among social
phenomena, and causal explanation is the
central form of social explanation. This
thesis cuts against those who argue that the
social sciences are intrinsically hermeneutic
and non-causal.
3
Manifesto
• Causal relations are not constituted by
regularities or laws connecting classes of
social events or phenomena. In fact, the
regularities of the social world fall far short
of the predictive and strict laws of nature
that characterize many domains of natural
phenomena.
4
Manifesto
• Social causal relations are constituted by the
causal powers of various social events,
conditions, structures, and the like, and the
singular causal mechanisms that lead from
antecedent conditions to outcomes.
Accordingly, a central goal of social
research is to identify the causal
mechanisms that give rise to social
outcomes.
5
Manifesto
• There is no such thing as pure social
causation from macro-state to macro-state;
instead, hypotheses about social causal
mechanisms must be constructed on the
basis of an account of the
“microfoundations” of the processes that are
postulated. Individuals choosing in the
context of structured circumstances of
choice are the engine of social change.
6
Manifesto
• Social entities and structures -- institutions,
ideologies, technological revolutions,
communications and transportation systems
-- all these exercise causal powers through
the effects that they have on individual
choices, preferences, beliefs, etc.
7
Manifesto
• Social scientists discover myriad causal
properties attaching to states, demographic
regimes, cropping practices, agricultural
technologies, transport systems, and the
like. Moreover, there is substantial causal
diversity in the mechanisms and properties
that social scientists discern.
8
Manifesto
• It is therefore methodologically defensible
to be eclectic in one’s causal hypothesizing;
there is no philosophical basis whatsoever
for supposing that there is a single unifying
social theory that could serve as the basis of
all social change.
9
II. Concepts of causation
• Causal regularities
– A causes B = events of type A are always followed by
events of type B
• Necessary and sufficient conditions
– A causes B = A is a necessary and/or sufficient
condition for B
• Causal mechanism
– A causes B = there exists a chain of causal mechanisms
leading from A to B
• Probabilistic causation
– A causes B = the occurrence of A raises the probability
10
of occurrence of B ( P(B|A)>P(B) )
Quantitative expressions
• Regression: O = aX + bY +cZ + e
• Conditional probability
• Correlation: history of smoking is correlated
with occurrence of lung cancer
• Y never occurs in the absence of X: X is a
necessary condition for Y.
• “Y would not have occurred if X had not
occurred.” Counterfactual reasoning.
11
Diagnosis
• The existence of causal mechanisms is
fundamental; the other criteria are symptoms of
there being a causal mechanism rather than
constitutive of the causal relation.
• If it is discovered that a social factor X appears to
be a cause of a social outcome Y, we need to
consider what the mechanism is that links X to Y
through the development and actions of ordinary
human individuals.
12
Causal realism
• In looking for a causal explanation of a particular outcome
P we are assuming that there is a set of properties,
conditions, and events which occurred prior to P and
which, as a consequence of the causal powers of these
factors, brought P into being.
13
Causal realism
• Causal realism postulates that there are real,
causally influential structures and processes which
have genuine historical effects and which are
amenable to rigorous scrutiny and explanation.
• Causal relations are not constituted by regularities
or laws.
• Rather, social causal relations are constituted by
the causal powers and causal mechanisms of
various social entities and circumstances.
14
Causal mechanisms
• Seek out the individual-level and local-level mechanisms
through which social outcomes emerge
• For example, “prisoners’ dilemmas,” “public goods
problems,” “principal-agent problem”
• Example: a system of paths through a forest
• Transport system as a mechanism of urbanization and
market extension
• University admissions system as a mechanism of
increasing / decreasing social stratification
15
People advocating causal realism
…
• Elster, Jon. 1989. Nuts and Bolts for the Social Sciences.
Cambridge: Cambridge University Press.
• Little, Daniel. 1991. Varieties of Social Explanation: An
Introduction to the Philosophy of Social Science. Boulder,
Colorado: Westview Press.
• McAdam, Doug, Sidney G. Tarrow, and Charles Tilly.
2001. Dynamics of Contention, Cambridge Studies in
Contentious Politics. New York: Cambridge University
Press.
16
People advocating causal realism
…
• Mahoney, James, and Dietrich Rueschemeyer. 2003.
Comparative Historical Analysis in the Social Sciences,
Cambridge Studies in Comparative Politics. Cambridge,
UK ; New York: Cambridge University Press.
• Hedström, Peter, and Richard Swedberg. 1998. Social
Mechanisms : An Analytical Approach to Social Theory,
Studies in Rationality and Social Change. Cambridge,
U.K. ; New York: Cambridge University Press.
17
Examples of social causal claims
• Population increase causes technological
innovation.
• A free press causes a low incidence of famine.
• The fiscal system of the ancien regime caused the
collapse of the French monarchy.
• Transport systems cause patterns of commerce and
habitation.
• New market conditions cause changes in systems
of norms
18
Examples
• A new irrigation system causes changes in family
organization
• Concentrated urban demand causes development
of an infrastructure to support a flow of timber and
grain into the metropolis
• The principal-agent problem represented by cattle
herding in Kenya causes the emergence of the
practice of bridewealth
• Citizens’ shared sense of justice causes stability or
instability of existing legal system
19
What gives rise to social causal
powers?
• Social entities exercise causal powers through the
effects that they have on individual choices,
preferences, and beliefs.
• Institutions, organizations, states, trading systems,
property systems – all have social effects through
their ability to influence individual’s choices.
• Institutions offer incentives, constraints, costs,
benefits, and opportunities.
• These factors have direct influence on individuals’
choices.
20
Causal powers of institutions
• The causal properties of a social entity consist in the
structures that it embodies that affect the actions of
individuals (through incentives, opportunities, powers,
information).
• Social entities exert influence in several possible ways:
–
–
–
–
–
–
They can alter incentives for individuals
They can alter preferences
They can alter cognitive and behavioral dispositions of individuals
They can alter beliefs
They can alter the powers or opportunities available to individuals.
They can impose costs on certain lines of action.
21
Example: race
• How does race influence outcomes such as educational
attainment?
• We can postulate a set of social arrangements that
influence the personality formation, belief formation,
preference formation, culture, and identity of young people
of color.
• We can postulate a system of opportunities and constraints
that affect persons of color differently from others.
• We can investigate the social psychology of race -- e.g.
“stereotype threat”.
• In other words, we can provide a micro-sociology of the
way that race and racism work in a particular society, and
trace out the effects these mechanisms have on individuals,
and the collective effects they have on behavior at the
22
group level.
• The most basic foundations for social causal
explanation are stories about the characteristics of
typical human agents within specific institutional
settings. The causal powers of a particular social
institution—a conscription system, a revenue
system, a system of democratic legislation—
derive from the incentives, powers, and
knowledge that these institutions provide for
participants. Social entities thus possess causal
powers in a derivative sense: they possess
characteristics that affect individuals’ behavior in
simple, widespread ways. Given features of the
common constitution and circumstances of
individuals, such alterations at the social level
produce regularities of behavior at the individual
level that eventuate in new social circumstances.
23
Agents and structures
• There are two directions of influence
between individuals and institutions within
the context of the microfoundations
framework.
• Structures constrain individuals.
• Individuals through their actions affect,
change, and invent institutions.
24
Generic causal properties of
social entities
• Generic causal properties of social institutions
derive from a common existential situation for a
group of agents; identify an accessible solution;
and infer that this institutional arrangement will
recur repeatedly.
• Generic social causal ascriptions thus depend on
common characteristics of agents (e.g. the central
axioms of rational choice theory).
25
IV. Causal reasoning in the social
sciences
• several distinct goals are possible:
– identify common processes; generalization
– identify singular or exceptional processes;
differentiation
• A common research goal: identify the
causal conditions that typically influence a
given outcome.
26
Small-N methods
• select a set of cases in which the variables of
interest are present (or absent); examine outcomes;
examine mechanisms and processes
• Probe the causal characteristics of these structures
through observation of their behavior in different
settings.
• Employ Mill’s methods of similarity and
difference; Boolean analysis of cases
• Complication: probabilistic causation
27
What is comparative inquiry?
• It is social or political research that focuses
on the causes and effects of social structures
and dynamics and pays close attention to
cross-case comparisons
• Identify similar structures and processes in
different social and historical settings
• Example: How does micro-organizational
structure affect the incidence of corruption?
28
Comparative methodology
• There are a small number of logical
methods of empirical inference through
which a causal hypothesis about categorical
variables may be tested.
• Mill’s methods have guided much
comparative historical sociology.
29
The causal diagram
30
The easy case: exceptionless
causation
• The requirements for this analysis:
– complete causal field--causal closure
– exceptionless causation--no probabilistic causes
• INUS conditions
31
Mill’s methods
• method of similarity
• method of difference
32
Mill’s methods generalized
• A truth table method for sorting out
necessary and sufficient conditions.
• The method begins with the exhaustive set
of conjuncts of independent conditions that
lead to the outcome.
33
Mill’s methods generalized: truth
table
• If we have an exhaustive list of potential
causes for an outcome; assurances that
causation is deterministic; and a lot of
patience, then we can construct a truth table
for all states of the variables (T or F; )
locate one case for each line of the table;
and derive a statement of the
necessary/sufficient conditions of
occurrence.
34
INUS conditions
• John Mackie’s analysis of categorical
causes
• INUS condition: insufficient but necessary
part of a condition which is itself
unnecessary but sufficient for the result
(Mackie 1974:62).
• BDEF v BDAF v BDAC v BDEC => R
• ((AvE)&B)&((CvF)&D)=> R
35
The hard case: probabilistic
causation
• Causal conditions exert influence probabilistically:
If Q1, Q2, … Qn, then there is a likelihood of P
that R will occur.
• Incomplete knowledge of the causal field;
possibility of unknown causes
• Mill’s methods are not useable.
• Causal relevance: A is causally relevant to the
occurrence of R iff P(R|A) <> P(R) (Salmon)
36
The hard case: probabilistic
causation
• Assume that the world is causally organized
according to the causal diagram. This
diagram involves six independent variables,
two intermediate causal variables (social
unrest and state crisis), and one final
variable (revolution). The states of variables
G and H are determined by the states of AF. And the state of R is determined by the
states of G and H.
37
Analysis
• With an arbitrarily large number of cases it
is possible to infer a great deal about the
structure of the causal diagram: relevance of
factors, independence or dependence of
factors, and conditional probability of
outcome given conjunctions of factors.
• The assumption of unlimited data is
unrealistic.
38
Case-study methods
• Causal realism gives a justification for the
application of a common method of social inquiry,
the case-study method.
• The case-study method is a legitimate method of
causal inquiry in the social sciences.
• The researcher may engage in an analysis that we
may call “process-tracing”: he/she needs to begin
to offer hypotheses about the causal relations
among the factors that emerge from the case study.
39
Examples of single-case studies
• The historical circumstances of the Chinese
Revolution.
• The outbreak of World War I.
• The occurrence of the Great Depression.
40
Large-N methods
• If a set of factors are causally related and if we can
produce a sufficient number of cases, then we
should be able to observe statistical associations
and correlations among these factors.
• We can therefore make use of statistical methods
in areas where it is possible to observe a
significant number of different cases, to suggest
causal hypotheses about relations among
variables.
41
The general question
• Suppose we are concerned about a feature that is
differentiated across a population: schooling
levels, income, occupation, test scores.
• Suppose we hypothesize that some socially
widespread conditions, properties, or influences
are causally relevant for explaining these
differences.
• How can we use observable data to evaluate
whether, and how strongly, a given factor is
causally influential?
42
Examples
• Example: to what extent is “propensity to migrate”
influenced by factors X,Y,Z? What are the causes
of high propensity to migrate?
• Example: are there features of institutional
organization that are more conducive to corruption
than other alternatives?
• Method: observe a large number of organizations;
code for a list of organizational characteristics;
“operationalize” the concept of corruption; and
look for correlations.
43
Concerns
• Spurious correlation; common cause
• Lieberson: selectivity, flaws of the assumption of “quasiexperiments”
• “Sociological ideas are best reintroduced into quantitative
sociological research by focusing on specifying the
mechanisms by which change is brought about in social
processes” (Sørensen 1998 : 264).
• “Developing theoretical ideas about social processes is to
specify some concept of what brings about a certain
outcome—a change in political regimes, a new job, an
increase in corporate performance, … The development of
the conceptualization of change amounts to proposing a
mechanism for a social process” (Sørensen 1998 : 23944
240).
Social theory and social
causation
• The empirical procedures commonly used
to probe causation in the social sciences
(Mill’s methods and its generalizations, and
various tests of statistical association)
almost always underdetermine the true
causal story for a given ensemble of
phenomena.
45
Social theory and social
causation
• Causal realism thus demands social
theory—collective action theory, theory of
bureaucracies and institutions, class conflict
theory, economic geography, rational choice
theory, theory of social-property regimes,
etc.—since we need to have an analysis of
the causal powers of the various factors in
order to account for the links in the causal
diagram.
46
47