No Job Name

Understanding Process Tracing

David Collier, University of California, Berkeley

ABSTRACT

Process tracing is a fundamental tool of qualitative analysis. This method is

often invoked by scholars who carry out within-case analysis based on qualitative data, yet

frequently it is neither adequately understood nor rigorously applied. This deﬁcit moti-

vates this article, which oﬀers a new framework for carrying out process tracing.The refor-

mulation integrates discussions of process tracing and causal-process observations, gives

greater attention to description as a key contribution, and emphasizes the causal sequence

in which process-tracing observations can be situated. In the current period of major inno-

vation in quantitative tools for causal inference, this reformulation is part of a wider, par-

allel eﬀort to achieve greater systematization of qualitative methods. A key point here is

that these methods can add inferential leverage that is often lacking in quantitative anal-

ysis. This article is accompanied by online teaching exercises, focused on four examples

from American politics, two from comparative politics, three from international relations,

and one from public health/epidemiology.

rocess tracing is a fundamental tool of qualitative

analysis. In the framework presented here,

it is

deﬁned as the systematic examination of diagnostic

evidence selected and analyzed in light of research

questions and hypotheses posed by the investigator.

Process tracing can contribute decisively both to describing polit-

ical and social phenomena and to evaluating causal claims. George

and Bennett have played the leading role in developing this

method as an essential form of within-case analysis,

and Fenno’s

“soaking and poking” is a kindred research procedure.

Although the idea of process tracing is often invoked by schol-

ars as they examine qualitative data, too often this tool is neither

well understood nor rigorously applied. Relatedly, the ﬁeld of qual-

itative methods in political science—in sharp contrast to quanti-

tative methods—is inadequately equipped with procedures for

teaching basic research tools, including process tracing.

This two-fold deﬁcit motivates this article, which oﬀers a new

framework for understanding, applying, and teaching process trac-

ing. The approach is distinctive in three ways.

ProcessTracing vis-à-vis CPOs.The evidence on which process trac-

ing focuses corresponds to what Collier, Brady, and Seawright

(2010a) call causal-process observations, or CPOs. The idea of

CPOs highlights the contrast between (a) the empirical foun-

dation of qualitative research, and (b) the data matrices ana-

lyzed by quantitative researchers, which may be called data-set

observations (DSOs). Some of the literature on which this arti-

cle draws (e.g., Brady 2010; Freedman 2010a; Mahoney 2010)

formulates arguments in terms of CPOs, rather than in terms

of process tracing per se. The present article treats these meth-

odological tools as two facets of the same research procedure.

Throughout, the article consistently refers to “process tracing”

to avoid applying two labels to what is basically the same

method.

Description. Careful description is a foundation of process tracing,

a perspective emphasized by Mahoney (2010, 125–31). Process

tracing inherently analyzes trajectories of change and causa-

tion, but the analysis fails if the phenomena observed at each

step in this trajectory are not adequately described. Hence, what

in a sense is “static” description is a crucial building block in

analyzing the processes being studied.

Sequence. Process tracing gives close attention to sequences of

independent, dependent, and intervening variables. Again, we

follow Mahoney, who has productively advanced this approach.

TEACHING EXERCISES

This new formulation of process tracing is accompanied by online

teaching exercises

that encompass diverse substantive areas.

a. American Politics: Fenno (1977) on members of Congress; Brady

(2010) on the 2000 Presidential election; Skocpol et al. (2000)

on civic associations; and Weaver (2007) on crime policy.

b. Comparative Politics: Lerner (1958) on social change in a Turk-

ish village; and Rogowski (2010) on the interaction of theory

and case studies.

c. International Relations: Tannenwald (1999) on the US “nuclear

taboo” after World War II; Bennett (2010) on the Fashoda cri-

sis, Germany’s expansion of military goals during World War I,

and Soviet nonintervention in Eastern Europe in 1989; and

Schultz (2001) on democracy and coercive diplomacy.

David Collier is Chancellor’s Professor of Political Science in the Department of

Political Science at the University of California, Berkeley. He can be reached at

[email protected].

The Teacher

.............................................................................................................................................................................................................................................................

doi:10.1017/S1049096511001429 PS • October 2011 823

PS: Political Science and Politics 44, No. 4 (2011): 823-30.

d. Public Health: Freedman (2010a) on major breakthroughs in

the history of epidemiology.

e. Detective Fiction: The Sherlock Holmes story “Silver Blaze”

(posted online with the exercises) serves as the basis for an

exercise, and also as a running example in parts of the follow-

ing presentation. This story is not social science, yet it provides

vivid illustrations of process tracing and is an engaging text for

teaching.

PROCESS TRACING, PRIOR KNOWLEDGE, AND DIAGNOSTIC

EVIDENCE

Process tracing, to reiterate, is an analytic tool for drawing descrip-

tive and causal inferences from diagnostic pieces of evidence—

often understood as part of a temporal sequence of events or

phenomena. Given the close engagement with cases and the cen-

trality of ﬁne-grained case knowledge, process tracing can make

decisive contributions to diverse research objectives, including:

(a) identifying novel political and social phenomena and system-

atically describing them; (b) evaluating prior explanatory hypoth-

eses, discovering new hypotheses, and assessing these new causal

claims; (c) gaining insight into causal mechanisms; and (d) pro-

viding an alternative means—compared with conventional regres-

sion analysis and inference based on statistical models—of

addressing challenging problems such as reciprocal causation,

spuriousness, and selection bias. Thus, qualitative tools can add

leverage in quantitative analysis. They can also strengthen causal

inference in small-N designs based on the matching and contrast-

ing of cases—designs which have great value, but whose contri-

bution to causal inference urgently needs to be supplemented by

within-case analysis.

Process tracing requires ﬁnding diagnostic evidence that pro-

vides the basis for descriptive and causal inference. How does the

researcher establish that a given piece of evidence is diagnostic?

Identifying evidence that can be interpreted as diagnostic

dependscentrallyonpriorknowledge.

Forthepurposeoftheonline

exercises, we distinguish four interrelated types of knowledge—

extending distinctions oﬀered by Waltz (1979), whose ideas are

important in the international relations examples included here.

Conceptual Frameworks. A ﬁrst type of prior knowledge involves

sets of interrelated concepts, often accompanied by general ideas

of how the concepts can be operationalized. These frameworks

thereby identify and link the topics seen as meriting analytic

attention. The framework often points to the counterfactuals

that conceptually establish what it means for a given phenom-

enon to be absent, that is, the “contrast space” (Garﬁnkel 1981)

that organizes the analysis.

Recurring Empirical Regularities. These are established patterns

in the relationships among two or more phenomena. Waltz

(1979, 1) states that this is “not simply . . . a relationship that

has been found, but . . . one that has been found repeatedly.”

The corresponding “if a, then b” (1979, 1) connection may be

viewed as causal, or it may be understood descriptively.

Theory-I.This builds on these recurring regularities by more tightly

connecting them as a set of insights into “a particular behavior

or phenomenon” (Waltz 1979, 2). Thus, many social scientists

seek to build theory “by collecting carefully veriﬁed, intercon-

nected hypotheses.”

Theory-II. A ﬁnal type of prior knowledge entails not only inter-

connected empirical regularities (Theory-I), but also a set of

statements that explain them, that is, oﬀering explanations of

why these regularities occur (Waltz 1979, 5). Theory-II may also

be called an explanatory model.

As is clear in the exercises, some studies are explicit and pre-

cise about the prior knowledge that frames the research, whereas

for other studies it is necessary to consult a wider literature to

understand the theoretical background. Unfortunately, as inves-

tigators write up their research, they may overstate the coher-

ence of the ﬁndings vis-à-vis prior knowledge—sometimes making

it hard to identify the theoretical starting point. Reconstructing

this starting point can require detective work—which is some-

times needed in evaluating diagnostic evidence in some of the

exercises.

Against this backdrop, we consider the contribution of pro-

cess tracing to descriptive and causal inference.

DESCRIPTIVE INFERENCE

Careful description is fundamental in all research, and causal

inference—whether assessed with qualitative or quantitative tools—

depends on it.

Close engagement with case knowledge in pro-

cess tracing can provide a good foundation for addressing this

task.

A key point must be underscored again. As a tool of causal

inference, process tracing focuses on the unfolding of events or

situations over time. Yet grasping this unfolding is impossible if

one cannot adequately describe an event or situation at one point

in time. Hence, the descriptive component of process tracing begins

not with observing change or sequence, but rather with taking

good snapshots at a series of speciﬁc moments. To characterize a

process, we must be able to characterize key steps in the process,

which in turn permits good analysis of change and sequence.

Mahoney (2010, 127–28) illustrates descriptive inference in pro-

cess tracing with Tannenwald’s (1999) study of the “Nuclear

Taboo.”

Tannenwald argues that the horriﬁed reaction to the

use of nuclear weapons at the end of the World War II created a

nuclear taboo that strongly inﬂuenced later US nuclear policy,

speciﬁcally decisions about the non-use of nuclear weapons dur-

ing subsequent military crises. Whereas this taboo grew out of

the reaction at the level of public opinion, it evolved into a nor-

mative mandate embraced by policy makers (1999, 462). A crucial

task in Tannenwald’s study is to establish empirically (a) that this

horriﬁed reaction did in fact occur; (b) how widespread it was;

and (c) that the elements of this reaction did indeed add up to a

nuclear taboo. Process tracing focuses on ﬁnding and interpret-

ing diagnostic evidence that addresses these descriptive tasks. This

nuclear taboo, in turn, is the key independent variable in the study

that is evaluated vis-à-vis rival explanations of the non-use of

nuclear weapons.

Lerner’s (1958) analysis of rapid “modernization” in a Turkish

village likewise illustrates the intensive description that should

be a foundation of process tracing.

This transformation results

from the election of a new national governing party and the sub-

sequent introduction of infrastructure that includes electricity and

a modern road to Ankara. The transformation of the village is the

dependent variable, and the author’s goal is to describe change in

this variable over time. The analysis focuses on dozens of speciﬁc

observations of social attributes and interactions; demographic

characteristics; and material objects, physical infrastructure, and

commercial establishments.

The Teacher: Understanding Process Tracing

824 PS • October 2011

These two examples of de-

scription diﬀer in important

ways. Tannenwald relies on di-

verse primary and secondary

sources—including oﬃcial doc-

uments, memoirs, and biogra-

phies—that shed light on the

politics of nuclear policy mak-

ing. By contrast, Lerner’s study

depends on intensive interview-

ing, carried out by his ﬁeld assis-

tants. Further, as noted, the

phenomenon described through

process tracing is Tannenwald’s

main independent variable; thus,

the nuclear taboo is her hypoth-

esized explanation for the US

post-World War II non-use of

nuclear weapons. By contrast,

for Lerner the posited modern-

ization of the village is the

dependent variable triggered by

an electoral shift and the initia-

tive of the victorious party to

build new infrastructure. Finally, while both Lerner and Tannen-

wald oﬀer a rich and detailed description of a key variable, Tan-

nenwald also gives substantial attention to rival explanations.

The two examples also illustrate another point: The qualita-

tive researcher should recognize that the ﬁne-grained descrip-

tion in process tracing sometimes relies on quantitative data.

This is certainly reasonable, given that—in the spirit of pursu-

ing multi-method research—the boundary between qualitative

and quantitative should not be rigid.

For Lerner, some of the

information is demographic, involving numerical data. As seen

in the exercises, Brady’s (2010) process tracing study employs

quantitative data on elections and voting. In parallel, Tannen-

wald could have assessed the pervasiveness of horriﬁed reactions

by counting their overall frequency, diﬀerent types of horriﬁed

reactions, and change in these counts over time. Process tracing

does indeed focus on single “nuggets” of information, yet some-

times this information involves counts and not just single actions

or occurrences.

A diﬀerent form of description, based on counterfactuals, is

illustrated by the Sherlock Holmes story “Silver Blaze.” Here the

central puzzle

is to explain the murder of John Straker, trainer

of the racehorse Silver Blaze. The focus is on a singular event that

cannot be disaggregated—a focus also common in process-tracing

research in international relations. With singular events, descrip-

tion may be based on comparison of the observed value of a given

variable with one or more hypothetical—i.e., counterfactual

—

values that are seen as plausible alternatives, but that do not occur

in the case being studied. The comparison depends on the con-

trast space noted above, which builds on the researcher’s back-

ground knowledge. Counterfactuals are important in diverse areas

of research (e.g., King, Keohane, and Verba 1994, 77–78, 88–89),

and they play a particularly visible role here.

CAUSAL INFERENCE

Basic ideas about applying process tracing to causal inference

may be summarized in terms of four empirical tests. Slightly

adapting the formulation of Bennett (2010), who builds on the

work of Van Evera (1997), the tests are classiﬁed according to

whether passing the test is necessary and/or suﬃcient for accept-

ing the inference. Based on these criteria, table 1 presents the

four tests: straw-in-the-wind, hoop, smoking-gun, and doubly deci-

sive. The table also notes the implications for rival hypotheses of

passing each test. If a given hypothesis passes a straw-in-the-

wind test, it only slightly weakens rival hypotheses; with hoop

tests it somewhat weakens them; with smoking-gun tests it sub-

stantially weakens them; and with doubly decisive tests passing

eliminates them—of course, with the usual caveat that the deﬁn-

itive elimination of a hypothesis is often hard to achieve in social

science.

Before we introduce causal inference, it is useful to reiterate

two ideas discussed above: process tracing can focus either on

recurring events or on a singular event; and although it is reason-

able to think of process tracing as a qualitative method, it some-

times relies on quantitative information.Three other points should

also be emphasized:

Speciﬁcation of Hypotheses. Careful, analytically informed speciﬁ-

cation of hypotheses is essential both in selecting and inter-

preting pieces of evidence, and in weighing them against one

another. Background knowledge is fundamental here.

Distinctions among Tests. The distinctions in table 1 are a use-

ful heuristic, but should not be taken rigidly. The decision to

treat a given piece of evidence as the basis for one of the

four tests can depend on the researcher’s prior knowledge,

the assumptions that underlie the study, and the speciﬁc

formulation of the hypothesis. Although in general the ap-

propriate test is clear, sometimes a piece of evidence treated

as a straw-in-the-wind might instead be viewed as the basis

for a hoop test or a smoking-gun test (see tables 4 and 5

below). Alternatively, it might simply be viewed as an “inter-

mediate” test, with corresponding implications for rival

hypotheses.

Table 1

Process Tracing Tests for Causal Inference

SUFFICIENT FOR AFFIRMING CAUSAL INFERENCE

No Yes

NECESSARY

FOR

AFFIRMING

CAUSAL

INFERENCE

Source:

Adapted from Bennett ~2010, 210!, who builds on categories formulated by Van Evera ~1997, 31–32!.

Yes

1. Straw-in-the-Wind 3. Smoking-Gun

a. Passing: Affirms relevance of hypothesis,

but does not confirm it.

a. Passing: Confirms hypothesis.

b. Failing: Hypothesis is not eliminated, but

is slightly weakened.

b. Failing: Hypothesis is not eliminated,

but is somewhat weakened.

c. Implications for rival hypotheses:

Passing

slightly

weakens them.

Failing

slightly

strengthens them.

c. Implications for rival hypotheses:

Passing

substantially

weakens them.

Failing

somewhat

strengthens them.

2. Hoop 4. Doubly Decisive

a. Passing: Affirms relevance of hypothesis,

but does not confirm it.

a. Passing: Confirms hypothesis and

eliminates others.

b. Failing: Eliminates hypothesis. b. Failing: Eliminates hypothesis.

c. Implications for rival hypotheses:

Passing

somewhat

weakens them.

Failing

somewhat

strengthens them.

c. Implications for rival hypotheses:

Passing

eliminates

them.

Failing

substantially

strengthens.

PS • October 2011 825

Assumptions and Interpretations. The decision about which test

is appropriate to a particular piece of evidence thus in-

volves diﬀerent assumptions and interpretations.

For

example, if researchers make the weaker assumption that a

given event (or other piece of evidence) may be a coincidence,

they should and will be more cautious. Alternatively, if

they make the stronger assumption—based on prior

knowledge—that it is probably not a coincidence, they may

arrive at a diﬀerent conclusion about accepting or rejecting

the hypothesis.

Against this backdrop, we discuss the four process-tracing tests

in table 1, using the Sherlock Holmes story “Silver Blaze” as a

running example. At this point it may be useful for readers to

examine the story itself.

Mapping the Holmes story onto the framework presented ear-

lier, we might say that the suspects are in eﬀect the hypotheses,

and the clues are causal process observations (CPOs). Table 2 pro-

vides an overview of the story and presents the hypotheses used

in illustrating the four tests, organized according to whether they

concern an independent, intervening, or dependent variable. As

indicated in the table, the mystery contains two causal puzzles:

explaining the murder of John Straker and the disappearance of

the horse. The following examples concentrate on the murder,

and references to the horse’s disappearance are considered when

crucial to the murder itself.

Straw-in-the-Wind Tests

These tests, illustrated in table 3, can increase the plausibility of

a given hypothesis or raise doubts about it, but are not decisive

by themselves. Straw-in-the-wind tests thus provide neither a

necessary nor a suﬃcient criterion for accepting or rejecting a

hypothesis, and they only slightly weaken rival hypotheses. Of

the four tests, these are the weakest and place the least demand

on the researcher’s knowledge and assumptions. Yet they pro-

vide valuable benchmarks in an investigation by giving an initial

assessment of a hypothesis. Furthermore, if a given hypothesis

passes multiple straw-in-the-

wind tests, it adds up to impor-

tant aﬃrmative evidence.

In “Silver Blaze,” one straw-

in-the-wind is based on the

clues about the bill for expen-

sive women’s clothing found in

Straker’s pocket and Straker’s

wife’s ignorance of the costly

dress that had been purchased.

This lends weight to Holmes’s

suspicion about Straker’s role

(H1) and to the idea that Straker

might have had a ﬁnancial

motive for throwing the race,

but is not by itself a decisive

piece of evidence. Another

straw-in-the-wind is one of the

most famous clues in all of

detective ﬁction: that the dog

presumably guarding the

horse’s stable “did nothing in

the night,” an observation that

points to the possibility that someone known to the dog—i.e.,

Straker—abducted the horse (H3). Yet it certainly does not con-

ﬁrm this hypothesis.

Hoop Tests

Hoop tests (table 4) set a more demanding standard. The hypoth-

esis must “jump through the hoop” to remain under consider-

ation, but passing the test does not by itself aﬃrm the hypothesis.

Although not yielding a suﬃcient criterion for accepting the expla-

nation, it establishes a necessary criterion. Hoop tests do not

conﬁrm a hypothesis, but they can eliminate it. Compared to the

straw-in-the-wind tests, passing hoop tests has stronger implica-

tions for rival hypotheses: it somewhat weakens their plausibility,

without precluding the possibility that alternative hypotheses may

be relevant.

Table 2

Overview of “Silver Blaze”

Causal Puzzle

To explain the murder of John Straker and, secondarily, the disappearance and whereabouts of the racehorse

Silver Blaze.

Main Characters

Silver Blaze

, the racehorse that is the favorite for the Essex Cup, has disappeared.

John Straker

, the horse’s trainer, has been killed by a terrible blow that shattered his head.

Fitzroy Simpson

, a prime suspect, has been lurking around the stable seeking inside information about the

race.

Ned Hunter

, a stable boy, has been drugged with opium concealed in curried mutton. He therefore fails to guard

Silver Blaze on the night of the horse’s disappearance.

Colonel Ross

is the owner of King’s Pyland Stables and of Silver Blaze.

Hypotheses

Independent Variables Intervening Variables Dependent Variables

H1. Romantic entanglement

started chain of events

H3. Straker abducted horse H6. Simpson killed Straker

H2. Chain of events started in

Straker household

H4. Straker planned to harm

horse

H7. Straker killed himself

H5. Straker practiced the injury

H8. Horse killed Straker

Table 3

Straw-in-the-Wind Tests

H1. Straker’s romantic entanglement set chain of events into motion.

Clues. A bill from an expensive women’s clothing store is found in Strak-

er’s pocket, and his wife is ignorant of the clothing in question.

Inference. The bill was owed by Straker for an expensive gift to another

woman, and Straker may have been in financial difficulty. This could

give him a motive for throwing the race.

Summary. This promising lead, a

straw-in-the-wind

, lends weight to H1,

but is not by itself a decisive piece of evidence.

H3. Straker abducted the horse.

Clue. The dog did nothing ~i.e., did not bark! in the night during which the

horse disappeared.

Inference. The person who approached the stable, possibly Straker, was

well-known to the dog. This raises questions about why Straker might

have gone to the stable. It suggests that perhaps he came to abduct

the horse, but does not strongly demonstrate this.

Summary. This

straw-in-the-wind

favors H3, but does not confirm it.

The Teacher: Understanding Process Tracing

826 PS • October 2011

Table 4 presents two hoop tests, both focused on the hypoth-

esis that Simpson killed Straker (H6). Simpson carried a poten-

tial murder weapon, so that he passes the corresponding hoop

test and is therefore not precluded as a suspect. However, his timid,

non-menacing appearance might seem to preclude his being a

murderer who “shattered” Straker’s head with a “savage” blow.

Hence, he fails this second hoop test. Table 4 also illustrates alter-

native interpretations of the same piece of evidence. One could

assume, as just stated, that a timid individual such as Simpson

would never commit a savage murder—thereby eliminating him

as a suspect through the hoop test. Alternatively, it may be unlikely,

but deﬁnitely not impossible, that a timid and nonmenacing per-

son would commit such a murder. This would yield a straw-in-

the-wind that casts doubt on the idea that that he is the murderer—

yet he remains a possible suspect.

Smoking-Gun Tests

The metaphor of a “smoking gun” conveys the idea that a suspect

who is caught holding a smoking gun is presumed guilty. How-

ever, those with no smoking gun may not be innocent. In other

words, this provides a suﬃcient but not necessary criterion for

accepting the causal inference. It can strongly support a given

hypothesis, but failure to pass does not reject it. If a given hypoth-

esis passes, it substantially weakens rival hypotheses.

In “Silver Blaze,” the ﬁrst smoking-gun test (table 5) is straight-

forward. The fact that the maid brought the curried mutton to

the stable shows that the initiative to drug the stable boy—a key

step in the chain of events—had to begin in Straker’s household

(H2). By contrast, the hypothesis that Straker planned to cause

harm (H4) is ambiguous and illustrates the importance of prior

knowledge and assumptions. Depending on such knowledge and

assumptions, the knife found with Straker can be viewed as

extraordinarily odd and suspicious, or only somewhat unusual.

Accordingly, the knife is alternatively a smoking-gun or a

straw-in-the-wind.

Doubly Decisive Tests

These tests provide strong inferential leverage that conﬁrms one

hypothesis and eliminates all others. They meet both the neces-

sary and suﬃcient standard for establishing causation. As Ben-

nett (2010, 211) notes, single tests that accomplish this are rare in

social science, but this leverage may be achieved by combining

multiple tests, which together support one explanation and elim-

inate all others.

Turning again to the Sherlock Holmes example (table 6), we

see that Simpson, Straker, and the horse are suspects in Straker’s

death. Simpson (H6) and Straker (H7) are removed from suspi-

cion by hoop tests. Based on straw-in-the-wind tests, one of which

might be interpreted as a smoking-gun test, Holmes infers that

the horse kicked Straker, thereby inﬂicting the grievous blow that

shattered his head. The conjunction of these diverse tests serves

to eliminate other suspects and establish the horse’s guilt (H8),

thereby meeting the standard of both necessity and suﬃciency.

Combining tests in this way poses important challenges. Cen-

tral here is Holmes’s method of elimination, a strategy invoked var-

ious times in Doyle’s stories. Put simply, when the investigator

has eliminated all plausible alternatives, the remaining scenario

must be the correct one. Variants of this method are widely recog-

nized, as with eliminative induction in Bayesian analysis (Vineberg

1996) and J.S. Mill’s (1974, 397–98) method of residues.

The method of elimination is especially relevant here because

although two suspects are deﬁnitely eliminated through hoop tests,

the guilt of the horse is established primarily on the basis of weaker

straw-in-the-wind tests. The procedure of elimination is valuable

because it relies centrally on the deﬁnitive exoneration of the ﬁrst

two suspects, and only secondarily on explicit demonstration of

the horse’s guilt.

Further, the method of elimination has special relevance to

the case of Silver Blaze, given that both Mill (1974, 398) and

Holmes emphasize the value of this method for discovering

unusual or even bizarre explanations—such as the kick of the horse

as a murder weapon (see table 6, H8, inference d). Mill states that

among his methods, “this is the most fertile in unexpected results”

(p. 398), and as Holmes puts it, “when you have eliminated the

impossible, whatever remains, however improbable, must be the

Table 4

Hoop Tests

H6. Simpson killed Straker.

Clue. Simpson had a potential murder weapon.

Inference. This weapon is consistent with the hypothesis, but does not

by itself demonstrate Simpson’s guilt.

Summary. Simpson had a potential weapon, so H6 passes this

hoop

test.

H6. Simpson killed Straker.

Clues. Simpson’s timid, non-menacing appearance, plus the fact that

Straker’s “head had been shattered by a savage blow from some heavy

weapon.”

Inference. With a stronger assumption based on his appearance, Simp-

son could not have inflicted the blow that shattered Straker’s head.

Alternative Inference. With a weaker assumption, Simpson’s appear-

ance raises doubts that he would have committed the murder, but does

not preclude it.

Summary. With a stronger assumption this is a

hoop

test which H6 fails;

with a weaker assumption it is a

straw-in-the-wind

test which casts

doubt on H6.

Table 5

Smoking-Gun Tests

H2. The chain of events started in Straker’s household.

Clues. The household maid brought the stable boy curried mutton, and

he was found later to have been drugged.

Inference. The curry was served to conceal the opium, which in turn was

used to drug the stable boy. When it is clear that the curry could only

have been introduced in the mutton by someone in Straker’s house-

hold, members of his household become inextricably linked to a key

causal step.

Summary. The clues yield a

smoking-gun

test that confirms H2.

H4. Straker planned to harm the horse.

Clue. Unusual, surgical knife found with Straker.

Inference: The knife is interpreted as

exceptionally

unusual—establishing

intent to harm.

Alternative Inference: The knife is interpreted as

somewhat

unusual,

suggesting, but hardly confirming, intent to harm. It might possibly be

a coincidence.

Summary. If the knife is exceptionally unusual, it is a

smoking gun

that

confirms H4. With a weaker interpretation that the knife was somewhat

unusual, it is a

straw-in-the-wind

that makes H4 more plausible, with

out confirming it.

PS • October 2011 827

truth. . . .”

In contemporary political science, a recurring con-

cern is that the discipline needs tools for discovering unexpected

and unusual explanations.The method of elimination merits atten-

tion as precisely this kind of tool.

Causal-Sequence Framework: Auxiliary Outcome Test

Another test is suggested by Mahoney’s (2010, 125–31) causal-

sequence framework. He introduces the idea of auxiliary out-

comes, which are not part of the main causal sequence yet provide

valuable inferential leverage.

Fruitful theories generate multi-

ple observable implications (King, Keohane, and Verba 1994,

passim), and a particular independent variable or mechanism

hypothesized to inﬂuence the dependent variable may also

aﬀect an auxiliary outcome. The inference thereby derived

may further support the causal importance of the independent

variable or intervening mechanism. In “Silver Blaze,” the lame

sheep are an example of an auxiliary outcome (table 7). Holmes

infers that Straker wished to practice the delicate operation

required to injure the horse (H5). This does not directly injure

the horse; it is a secondary outcome that makes Holmes’s ideas

about the central causal process more plausible. Auxiliary out-

come tests generally yield straws-in-the-wind, as occurs in this

example.

Mahoney further illustrates the auxiliary outcome test with a

social science example: Luebbert’s (1991) famous book, Liberal-

ism, Fascism, or Social Democracy. Luebbert’s central argument is

that a “red-green” coalition of socialist parties and the middle

peasantry was a key factor in the formation of national-political

economies in interwar Europe. Mahoney shows that while this

claim is partly developed through small-N comparative research

and partly through a focus on mechanisms, Luebbert also builds

his case by arguing that if a red-green alliance really did foster

social democracy, it should have left behind other markers, includ-

ing the reluctance of socialists to challenge the distribution of

wealth in the countryside (Mahoney 2010, 130). The discovery of

such auxiliary outcomes suggests that the red-green alliance had

a key impact on other domains of national politics. This ﬁnding

reinforces the idea that the alliance was highly inﬂuential, yield-

ing stronger grounds for inferring that it also shaped the national

political-economic regime (Mahoney 2010, 130).

CONCLUSION

This article seeks to improve the practice of process tracing as a

strategy of qualitative analysis, a strategy that can also contribute

to quantitative research. The discussion is accompanied by the

online exercises focused on ten empirical studies, from diverse

subﬁelds, aimed at encouraging careful thinking—and productive

teaching—about process tracing.

Three concluding points merit emphasis. First, as Brady, Col-

lier, and Seawright (2010, 22) note, “both qualitative and quanti-

tative research are hard to do well.” Qualitative tools such as

process tracing can address some challenges faced in quantitative

analysis, but process tracing faces serious issues in its own right.

Doubts may arise as to which causal-inference test is appropriate.

The analysis may face standard problems of missing variables.

Measurement error can be an issue, and probabilistic relation-

ships are harder to address than in quantitative research. This

article is intended as one step in developing and reﬁning tools for

process tracing—and it is urgent that it not be the last step. More

work must be done.

Second, in a given study, how does one begin to carry out pro-

cess tracing? It is certainly valuable to approach process tracing

with the expectation of using the causal inference tests presented

in table 1, yet these tests are not always easy to apply. It can there-

fore be productive to start with a good narrative or with a timeline

that lists the sequence of events. One can then explore the causal

ideas embedded in the narratives, consider the kinds of evidence

Table 6

Building a Doubly Decisive Test by

Evaluating Alternative Hypotheses

H6. Simpson killed Straker.

Clues. The household maid brought the stable boy curried mutton, and

he was found later to have been drugged.

Inference. The curry served to conceal the opium, which in turn was

used to drug the stable boy. Because the curry could only have been

introduced in the mutton by someone in Straker’s household, Simpson

is excluded from consideration.

Summary for H6. This

hoop

test rejects H6, eliminating Simpson as a

suspect.

H7. Straker killed himself.

Clue. Straker’s head was “shattered by a savage blow from some heavy

weapon.”

Inference. Straker could not have shattered his own head with such a

blow.

Summary for H7. This

hoop

test rejects H7, eliminating Straker as a

suspect.

H8. The horse killed Straker.

Clues. ~a! Bill in Straker’s pocket and wife’s ignorance of dress; ~b! Strak-

er’s unusual surgical knife; ~c! matches and candle, and the horse dis-

appeared at night; ~d! the unusual form of death—Straker’s head was

“shattered by a savage blow from some heavy weapon.”

Inferences. ~a! Straker’s romantic entanglement and resulting financial

difficulty provided a motive to throw the race ~

straw-in-the-wind

, table

3!; ~b! Straker planned to harm the horse ~

smoking gun

straw-in-the-

wind

, table 5!; ~c! when Straker attempted to harm the horse at night,

the candle and prick of the knife frightened the horse and led to a fatal

kick ~

straw-in-the-wind

!; ~d! the unusual form of death ~savage blow!

points to an extreme or unusual cause, i.e., the horse’s kick ~

straw-in-

the-wind

Summary for H8. The combined weight of four

straws-in-the-wind

, one

of which may be a

smoking gun

, strongly favor the hypothesis that the

horse killed Straker.

Summary of Doubly Decisive Test. Two of the three suspects are elim-

inated by

hoop

tests, leaving only the horse. A series of additional infer-

ences strongly implicate the horse. The combined weight of evidence

confirms H8, that the horse killed Straker.

Table 7

Auxiliary Outcome

H5. Straker practiced in preparation for injuring the horse.

Clue. Lame sheep.

Inference. Straker used the sheep to practice a delicate operation with

his surgical knife—which he planned to use for inflicting an undetect-

able injury to the horse. The inference is not that the sheep’s lameness

is a step in the central explanatory chain; rather lends further support

to Holmes’s understanding of that chain

Summary. The lameness is a

straw-in-the-wind

that favors H5 without

confirming it.

The Teacher: Understanding Process Tracing

828 PS • October 2011

that may conﬁrm or disconﬁrm these ideas, and identify the tests

appropriate for evaluating this evidence.

Finally, along with the value per se of reﬁning process tracing,

this discussion is important in wider debates on political meth-

odology. Political science is in a period of major innovation in

reﬁning tools for quantitative analysis, and in particular, quanti-

tative tools for causal inference. This trend has produced some

worries among qualitative researchers about the adequacy of their

own tools, and perhaps it has intensiﬁed the skepticism of some

quantitative researchers about causal inference in qualitative stud-

ies. This skepticism led the eminent statistician David Freedman

(2010a) to counter with the argument that the kind of qualitative

analysis involved in process tracing is indeed a type of scientiﬁc

inquiry in its own right. In that spirit, the goal here is to take

steps toward placing this form of inquiry on a more rigorous

foundation.

䡲

NOTES

Among the several colleagues who provided valuable comments on this article, Maria

Gould and three anonymous reviewers for PS deserve special thanks.

1. The approach discussed here diﬀers from other research traditions that can be

linked to the idea of process tracing—for example, the work on mechanisms of

Tilly (2001) and McAdam, Tarrow, and Tilly (2001).

2. Within-case analysis can become multi-case analysis if diﬀerent facets of the

initial “case” are analyzed. The key idea here is that the point of departure is a

single case, when viewed from the perspective of a wider comparative analysis

focused on a larger N.

3. George (1979); George and McKeown (1985); George and Bennett (2005);

Bennett (2008, 2010). On “soaking and poking,” see Fenno (1977, 884; 1978,

xiv; 1998, v). Process tracing has much in common with Lazarsfeld’s (1940,

preface) procedure of “discerning”; Campbell’s (1975, 181–82) “pattern match-

ing,” which is also advocated by Yin (1984/2008); Sewell’s (1996, 261) “causal

narrative”; Bates et al.’s (1998) “analytic narratives”; and Hall’s (2003, 391–95)

“systematic process analysis.”

4. See ^tinyurl.com/DavidCollier&.

5. Ideas about these designs based on a matching of cases are often drawn from

J.S. Mill (1974) and Przeworski and Teune (1970). For a comment on the

weakness of these designs for causal inference, see Collier, Brady, and Sea-

wright (2010a, 10).

6. Addressing this question raises issues about the logic of inquiry and the form

of social scientiﬁc knowledge that are well beyond the scope of this discus-

sion. Only a few basic points are addressed here that are salient for the accom-

panying exercises.

7. Obviously, such prior knowledge is essential in all research, both qualitative

and quantitative.

8. Waltz calls claims about these regularities “law-like statements” (p. 1). We

prefer the alternative label used here.

9. The expressions “descriptive inference” and “causal inference” are employed

here in the sense of King, Keohane, and Verba (1994, 7–8, chaps. 2–3). Their

usage can be seen as approximating an ordinary language meaning of “de-

scription” and “causation”; and by “inference” they mean that researchers

have “the goal of making inferences that go beyond the particular observa-

tions collected,” that is, they are analyzed within the larger framework used

by the investigator. This usage contrasts with ideas of “descriptive inference”

and “statistical inference” that are standard in the work of statisticians (e.g.,

Berk 2004, chap. 11).

10. Achieving good description in this sense, and developing fruitful ideas about

the unfolding of the process, may of course interact in an iterative manner.

11. Tannenwald’s study is also discussed in Collier, Brady, and Seawright (2010a,

189–90; 2010b, 509).

12. Lerner’s analysis—which is the focus of one of the exercises—is closely tied to

modernization theory, which might concern some readers; and at certain

points the presentation seems condescending. Further, the female interviewer

is presented in a sexist way (although in survey research, selecting interview-

ers in light of characteristics such as these is widely recognized as important).

However, these drawbacks are outweighed by the opportunity presented by

the chapter to illustrate the practice of making careful observations, and also

to see how they can be integrated into a complex picture of social change.

13. For a framing of qualitative vis-à-vis quantitative in terms of four dimensions,

see Collier, Brady, and Seawright (2010a, 177–82).

14. Another puzzle is explaining the disappearance of the horse, but as Holmes

himself emphasizes, that is a secondary issue (see p. 11 in the accompanying

online version of the story).

15. See Levy’s (2008) excellent discussion of counterfactuals and case studies.

16. There is a parallel here to the idea in statistical work that the test does not

stand on its own, but rather is shaped by prior assumptions. In quantitative

analysis, the construction of the statistical model depends heavily on such

assumptions, and in general the statistical test does not directly evaluate these

assumptions. Rather, it estimates the relationship based on the supposition

that the model assumptions, as well as the underlying assumption of causal-

ity, are true. See, for example, Freedman (2010b).

17. In one story, Sherlock Holmes takes a strong stand on coincidences (“Adven-

ture of the Second Stain”; in Doyle 1960, this is on p. 655). Watson refers to

the juxtaposition of two key events as “an amazing coincidence.” Holmes

replies: “A coincidence! The odds are enormous against its being a coinci-

dence. No ﬁgures can express them. No, my dear Watson, the two events are

connected—must be connected. It is for us to ﬁnd the connection.” Ironically,

it turns out that these two events are only tangentially connected, so Watson’s

statement was closer to the truth than Holmes’s, and the weaker assumption

was more appropriate. In another story, Holmes is initially more cautious

about inferences and coincidences, but then he backtracks and insists on the

certitude of his inferences (“The Sign of Four,” chap. 1; in Doyle 1960, this is

on p. 93).

18. “The Sign of Four,” chap. 6 (in Doyle 1960, this is on p. 111).

19. A further perspective on unusual or bizarre explanations (see again table 6,

H8, inference d) derives from William James’s famous dictum that “every

diﬀerence must make a diﬀerence.” To put this in a less extreme form, it

might be said that some diﬀerences make a diﬀerence. In this instance, the

form of the murder was so distinctive that it called for a distinctive

explanation—which turned out to be the kick of a horse. On William James,

see Copi (1953, 331–32).

20. Andrew Bennett (personal communication) has underscored the parallel here

with diagnostic tests in medicine.

21. Waldner (2011) oﬀers an interesting discussion of such issues.

REFERENCES

Bates, Robert H., Avner Greif, Margaret Levi, Jean-Laurent Rosenthal, and Barry

Weingast. 1998. Analytic Narratives. Princeton, NJ: Princeton University Press.

Bennett, Andrew. 2008. “Process Tracing: A Bayesian Perspective.” In The Oxford

Handbook of Political Methodology, ed. Janet M. Box-Steﬀensmeier, Henry E.

Brady, and David Collier, 702–21. New York: Oxford University Press.

Bennett, Andrew. 2010. “Process Tracing and Causal Inference.” In Rethinking

Social Inquiry: Diverse Tools, Shared Standards, 2nd ed., ed. Henry E. Brady and

David Collier, 207–19. Lanham, MD. Rowman and Littleﬁeld.

Berk, Richard A. 2004. Regression Analysis: A Constructive Critique. Newbury Park,

CA: Sage.

Brady, Henry E. 2010. “Data-Set Observations versus Causal-Process Observations:

The 2000 U.S. Presidential Election.” In Rethinking Social Inquiry: Diverse Tools,

Shared Standards, 2nd ed., ed. Henry E. Brady and David Collier, 237–42. Lan-

ham, MD: Rowman and Littleﬁeld.

Brady, Henry E., David Collier, and Jason Seawright. 2010. “Refocusing the Discus-

sion of Methodology.” In Rethinking Social Inquiry: Diverse Tools, Shared Stan-

dards, 2nd ed., ed. Henry E. Brady and David Collier, 15–31. Lanham, MD:

Rowman and Littleﬁeld.

Campbell, Donald T. 1975. “‘Degrees of Freedom’ and the Case Study.” Compara-

tive Political Studies 8 (2) : 178–93.

Collier, David, Henry E. Brady, and Jason Seawright. 2010a. “Sources of Leverage

in Causal Inference: Toward an Alternative View of Methodology.” In Rethink-

ing Social Inquiry: Diverse Tools, Shared Standards, 2nd ed., ed. Henry E. Brady

and David Collier, 161–99. Lanham, MD: Rowman and Littleﬁeld.

Collier, David, Henry E. Brady, and Jason Seawright. 2010b. “Outdated Views of

Qualitative Methods: Time to Move On.” Political Analysis 18 (4) : 506–13.

Copi, Irving M. 1953. Introduction to Logic. New York: Macmillan.

Doyle, Arthur Conan. 1960/1986. The Complete Sherlock Holmes, Vol. 1 and 2. New

York: Doubleday. The 1986 edition (New York: Bantam Classics) is available in

paperback, and all the stories are readily available online.

Fenno, Richard F., Jr. 1977. “U.S. House Members in Their Constituencies: An

Exploration.” American Political Science Review 71 (3): 883–917.

Fenno, Richard F., Jr. 1978. Home Style: House Members in their Districts. Boston:

Little, Brown.

Fenno, Richard F., Jr. 1998. Senators on the Campaign Trail: The Politics of Represen-

tation. Norman: University of Oklahoma Press.

PS • October 2011 829

Freedman, David A. 2010a. “On Types of Scientiﬁc Inquiry: The Role of Qualita-

tive Reasoning.” In Rethinking Social Inquiry: Diverse Tools, Shared Standards,

2nd ed., ed. Henry E. Brady and David Collier, 221–36. Lanham, MD: Rowman

and Littleﬁeld.

Freedman, David A. 2010b. Statistical Models and Causal Inference: A Dialogue with

the Social Sciences. New York: Cambridge University Press.

Garﬁnkel, Harold. 1981. Forms of Explanation. Rethinking the Questions in Social

Theory. New Haven, CT: Yale University Press.

George, Alexander L. 1979. “Case Studies and Theory Development: The Method

of Structured, Focused Comparison.” In Diplomacy: New Approaches in History,

Theory and Policy, ed. Paul Gordon Lauren, 43–68. New York: The Free Press.

George, Alexander L., and Andrew Bennett. 2005. Case Studies and Theory Develop-

ment in the Social Sciences. Cambridge, MA: MIT Press.

George, Alexander L., and Timothy J. McKeown. 1985. “Case Studies and Theories

of Organizational Decision Making.” In Advances in Information Processing in

Organizations, Vol. 2. ed. Robert F. Coulam and Richard A. Smith, 21–58.

Greenwich, CT: JAI Press.

Hall, Peter A. 2003. “Aligning Ontology and Methodology in Comparative Re-

search.” In Comparative Historical Analysis in the Social Sciences, ed. James Ma-

honey and Dietrich Rueschemeyer, 373–404. New York: Cambridge University

Press.

King, Gary, Robert O. Keohane, and Sidney Verba. 1994. Designing Social Inquiry:

Scientiﬁc Inference in Qualitative Research. Princeton, NJ: Princeton University

Press.

Lazarsfeld, Paul F. 1940. “Introduction.” In The Unemployed Man and His Family,

Mirra Komarovsky, ix–xii. New York: Dryden.

Lerner, Daniel. 1958. The Passing of Traditional Society: Modernizing the Middle East.

New York: The Free Press. A somewhat abridged version, which unfortunately

omits some key details, is reprinted in Development: A Cultural Studies Reader,

ed. Susanne Schech and Jane Haggis (Oxford: Blackwell, 2002). This abridged

version is available online in Google Books at: http://tinyurl.com/Lerner-

Grocer-Chief.

Levy, Jack S. 2008. “Counterfactuals and Case Studies.” In The Oxford Handbook of

Political Methodology, ed. Janet M. Box-Steﬀensmeier, Henry E. Brady, and

David Collier, 627–44. New York: Oxford University Press.

Luebbert, Gregory M. 1991. Liberalism, Fascism, or Social Democracy: Social Classes

and the Political Origins of Regimes in Interwar Europe. New York: Oxford Univer-

sity Press.

Mahoney, James. 2010. “After KKV: The New Methodology of Qualitative Re-

search.” World Politics 62 (1): 120–47.

McAdam, Doug, Sidney Tarrow, and Charles Tilly. 2001. Dynamics of Contention.

New York: Cambridge University Press.

Mill, John Stuart. 1974. A System of Logic. Toronto: University of Toronto Press.

Przeworski, Adam, and Henry E. Teune. 1970. The Logic of Comparative Social In-

quiry. New York: Wiley.

Rogowski, Ronald. 2010. “How Inference in the Social (but Not the Physical) Sci-

ences Neglects Theoretical Anomaly.” In Rethinking Social Inquiry: Diverse

Tools, Shared Standards, 2nd ed., ed. Henry E. Brady and David Collier, 89–97.

Lanham, MD: Rowman and Littleﬁeld.

Schultz, Kenneth A. 2001. Democracy and Coercive Diplomacy. New York: Cam-

bridge University Press.

Sewell, William. 1996. “Three Temporalities: Toward an Eventful Sociology.” In

The Historic Turn in the Human Sciences, ed. Terrence J. McDonald, 245–80. Ann

Arbor: University of Michigan Press.

Skocpol, Theda, Herbert Gans, and Ziad Munson. 2000. “A Nation of Organizers:

The Institutional Origins of American Civic Life.” American Political Science

Review 94 (3) : 527–46.

Tannenwald, Nina. 1999. “The Nuclear Taboo: The United States and the Norma-

tive Basis of Nuclear Non-Use.” International Organization 53 (3): 433–68.

Tilly, Charles. 2001. “Mechanisms in Political Processes.” Annual Review of Political

Science 4: 21–41.

Van Evera, Stephen. 1997. Guide to Methods for Students of Political Science. Ithaca,

NY: Cornell University Press.

Vineberg, Susan. 1996. “Eliminative Induction and Bayesian Conﬁrmation

Theory.” Canadian Journal of Philosophy 26 (June): 257–66.

Waldner, David. 2011. “Process Tracing: Its Promise and Its Problems.” Paper

presented at the Institute for Qualitative and Multi-Method Research, Syra-

cuse University.

Waltz, Kenneth N. 1979. Theory of International Politics. New York: McGraw-Hill.

Weaver, Vesla M. 2007. “Frontlash: Race and the Development of Punitive Crime

Policy.” Studies in American Political Development 21 (Fall): 230–65.

Yin, Robert K. 1984/2008. Case Study Research: Design and Methods, 1st and 4th ed.

Beverly Hills, CA: Sage.

The Teacher: Understanding Process Tracing

830 PS • October 2011