Process Tracing

Process tracing is a within-case method for investigating causal mechanisms. Rather than asking whether X causes Y (which cross-case comparison can address), process tracing asks how X causes Y by unpacking the steps that connect a cause to an outcome. It is one of the most widely used qualitative methods in political science and international relations, and it pairs naturally with case study research.

What Is Process Tracing?

Process tracing examines the chain of events, decisions, and actions that link a proposed cause to an observed outcome within a single case. The goal is to identify and test a causal mechanism — the specific sequence of steps through which a cause produces an effect.

As Collier (2011) puts it, process tracing is “the systematic examination of diagnostic evidence selected and analyzed in light of research questions and hypotheses posed by the investigator.” You are not simply narrating what happened. You are testing whether the evidence is consistent with a proposed causal chain and inconsistent with alternatives.

Within-case vs. cross-case: Process tracing operates within a single case, examining the internal causal logic in detail. This distinguishes it from comparative methods (such as MSSD or MDSD), which examine patterns across cases. Many strong theses combine both approaches — using cross-case comparison to identify which variables matter and process tracing to explain how they work. See Related Methods for more on this connection.

Key Concepts

Causal mechanism. The central object of process tracing. A causal mechanism is a theorized sequence of steps (or entities and activities) that transmits a causal force from X to Y. For example, if you hypothesize that economic sanctions cause policy change, the mechanism might run: sanctions imposed –> economic costs increase –> key domestic elites lose revenue –> elites pressure the government –> government concedes. Each step must be specified in advance and tested against evidence.

Diagnostic evidence. Process tracing relies on evidence that can discriminate between competing explanations. Not all evidence is equally useful. The key question is: does this piece of evidence confirm or disconfirm a specific step in the proposed mechanism? Beach and Pedersen (2019) distinguish between different types of observable manifestations that can serve as diagnostic evidence for each step of a mechanism.

Variants of process tracing. Beach and Pedersen (2019) identify three main variants: theory-testing, in which you start with a hypothesized mechanism derived from the literature and test whether it operated in your case; theory-building, in which you work backwards from an observed outcome to identify the mechanism that produced it; and explaining-outcome, in which you construct a case-specific explanation that may combine multiple mechanisms. This guide focuses primarily on the theory-testing variant because it is the most structured and the most common in student theses. If you think theory-building or explaining-outcome better suits your research question, consult Beach and Pedersen (2019, chs. 3–5) and discuss the choice with your supervisor.

Equifinality. Multiple different causal paths can produce the same outcome. Process tracing does not prove that your mechanism is the only possible explanation — it demonstrates that the mechanism operated in your case. You must always consider and address alternative mechanisms that could also explain the outcome.

Think of it like a detective investigating a crime. A detective does not just observe that a suspect was at the scene (correlation). They trace the sequence: motive, means, opportunity, forensic evidence, witness testimony. Each piece of evidence either strengthens or weakens the case. Process tracing works the same way — you build a chain of evidence linking cause to outcome, step by step.

Types of Evidence Tests

Mahoney (2012) and Collier (2011) identify four classic tests for evaluating evidence in process tracing. Each test has different implications for confirming or eliminating a hypothesis:

Test	Passing confirms?	Failing eliminates?	Logic
Straw-in-the-wind	Slightly	No	Consistent but not conclusive; neither confirms nor eliminates
Hoop	No	Yes	A necessary condition; failing it eliminates the hypothesis
Smoking gun	Yes	No	Strongly confirms if found, but absence does not eliminate
Doubly decisive	Yes	Yes	Both necessary and sufficient; rare in practice

Straw-in-the-wind tests offer weak evidence. Finding that a government official mentioned sanctions in a speech is consistent with the sanctions mechanism, but it does not prove the mechanism operated. Failing the test does not eliminate it either.

Hoop tests set minimum thresholds. If you claim sanctions caused economic pain to elites, there must be evidence of actual economic losses. If there is no such evidence, the mechanism fails. But finding economic losses alone does not confirm the full mechanism.

Smoking gun tests provide strong confirmation. A leaked memo showing that business elites explicitly pressured the government to change policy because of sanctions would be powerful evidence. However, not finding such a memo does not mean the pressure never occurred.

Doubly decisive tests are both necessary and sufficient — the gold standard, but extremely rare in social science.

Ask yourself

For your proposed mechanism, which steps could you subject to a hoop test? What evidence would need to exist for each step to remain plausible?
Can you think of a smoking gun that would strongly confirm one of your causal steps?
What alternative mechanisms could explain the same outcome, and what evidence would distinguish your explanation from those alternatives?

When to Use It

Process tracing is appropriate when:

Your research question asks how or why an outcome occurred in a specific case
You have a proposed causal mechanism you want to test or develop
You have (or can obtain) detailed evidence about the sequence of events within the case — such as documents, interviews, speeches, policy records, or media reports
You want to go beyond correlation to examine the causal logic connecting variables

Process tracing is less appropriate when:

Your question is primarily about how widespread a pattern is across many cases (use cross-case comparison instead)
You lack access to detailed within-case evidence
You are interested in measurement or description rather than causal explanation

BA vs. MA expectations: BA students are expected to specify a plausible causal mechanism in advance, identify the key steps, and assess the evidence for each step in a structured way. MA students should engage more deeply with the methodological literature, justify their choice of process tracing variant, apply the evidence tests explicitly, and demonstrate greater sophistication in handling alternative explanations and equifinality. See Assessment Standards for how methodology is evaluated.

How to Apply It

Process tracing involves a structured sequence of steps. The sections below walk through each one.

Step 1: Specify Your Causal Mechanism in Advance

Before examining your case, you must articulate the mechanism you expect to find. This means breaking your proposed explanation into discrete, sequential steps. Each step should identify who does what and why.

Write out the full chain: X –> Step 1 –> Step 2 –> Step 3 –> Y. Be specific. “Economic pressure leads to policy change” is not a mechanism. “Sanctions reduce export revenue –> firms linked to the ruling coalition lose profits –> business elites lobby the president for concessions –> the president calculates that the cost of maintaining the policy exceeds the cost of conceding” is a mechanism.

Draw it out. Create a diagram of your mechanism with each step in a separate box and arrows connecting them. This forces you to be explicit about the causal logic and helps you identify where the chain might break down.

Step 2: Derive Observable Implications for Each Step

For each step in the mechanism, ask: if this step actually occurred, what evidence should I expect to find? And what evidence would be inconsistent with it?

These observable implications are your empirical predictions. For example, if you claim that business elites pressured the government, you would look for evidence such as meeting records, public statements by business leaders, lobbying disclosures, media reports of business opposition, or shifts in political donations.

Think about what types of evidence tests (hoop, smoking gun, straw-in-the-wind) each piece of evidence constitutes. Prioritize collecting evidence that can serve as hoop or smoking gun tests, since these have the most inferential power.

Step 3: Collect and Evaluate Evidence Systematically

Gather evidence for each step of the mechanism. This typically involves primary sources (government documents, speeches, archival records, interview transcripts, media coverage) and may be supplemented by secondary scholarship.

For each piece of evidence, assess:

What step does it speak to? Map every piece of evidence to a specific step in your mechanism.
What kind of test does it represent? Is it a hoop test (necessary but not sufficient), a smoking gun (sufficient but not necessary), or a straw-in-the-wind?
How reliable is the source? Consider the source’s potential biases, the context in which it was produced, and whether it can be corroborated.
Does it support or weaken the mechanism? Be honest. If evidence contradicts a step, report it and assess what it means for the overall argument.

Step 4: Assess the Overall Chain

After evaluating evidence for each step, make a judgment about the mechanism as a whole. Did every step pass at least a hoop test? Were there any smoking guns? Are there steps with weak or missing evidence?

A strong process tracing analysis does not require a smoking gun for every step. But it does require that no step fails a hoop test (which would break the chain) and that the overall body of evidence is more consistent with your mechanism than with plausible alternatives.

Step 5: Test Alternative Explanations

Identify at least one or two alternative mechanisms that could explain the same outcome. Subject them to the same evidence tests. If the evidence is equally consistent with an alternative mechanism, you cannot claim strong support for your preferred explanation.

This step is not optional. One of the most common weaknesses in student theses is confirming only the preferred explanation without seriously testing alternatives. Bennett and Checkel (2015) emphasize that the inferential power of process tracing depends on actively engaging with competing hypotheses.

Exercise: Draft your causal mechanism

State your research question and the outcome you want to explain.
Write out the proposed causal mechanism as a step-by-step chain (aim for 3-5 steps).
For each step, list at least two pieces of observable evidence you would expect to find if the step occurred.
Identify the strongest alternative explanation for your outcome and note how you would distinguish it from your preferred mechanism.
Bring this to your supervisor for feedback before you begin detailed evidence collection.

Structuring Your Thesis

Process tracing maps directly onto a standard thesis structure, but with specific requirements for each chapter. Here is how to organize the key sections.

Introduction

Present your research question (a how or why question about a specific case), motivate the puzzle, and briefly preview the proposed mechanism. State clearly that you are using process tracing as your method.

Literature Review

Review the scholarly debate around your topic. Identify the competing explanations in the literature for the type of outcome you are studying. These competing explanations become the basis for the alternative mechanisms you will test.

Analytical Framework

This chapter is critical for process tracing theses. It must accomplish three things:

Explain and justify your method. Define process tracing, explain why it is appropriate for your research question, and cite the methodological literature (e.g., Beach and Pedersen 2019; Bennett and Checkel 2015).
Specify the proposed causal mechanism. Lay out the full mechanism step by step, explaining the theoretical logic behind each step. This is where you derive the mechanism from your theory — not from your data.
State the observable implications. For each step, identify what evidence you expect to find and what types of evidence tests you will apply. Also specify the alternative mechanisms you will test and their observable implications.

The mechanism must appear before the findings. A common mistake is to present the mechanism only after describing the evidence, which makes the analysis look like a post hoc narrative rather than a genuine test. Specify the mechanism in the analytical framework chapter so that the reader knows what you are testing before you present the results.

Findings

Organize the findings chapters around the steps of the mechanism, not around a chronological narrative. For each step:

Present the relevant evidence
Assess what type of test the evidence represents
State whether the step passes or fails

If your mechanism has three to five steps, you might dedicate a subsection to each step within a single findings chapter, or split the analysis across two chapters for longer mechanisms. The key is that the structure follows the mechanism, not the timeline.

Address alternative explanations within the findings chapters or in a dedicated section. Show that the evidence is more consistent with your mechanism than with alternatives.

Conclusion

The conclusion evaluates the overall causal chain. Did the mechanism hold? Which steps were most strongly supported? Where was the evidence weakest? What does this tell you about the broader theoretical debate you engaged in your literature review? Discuss limitations honestly — particularly any steps where evidence was thin — and suggest directions for future research.

Example structure for a process tracing thesis:

Introduction
Literature Review
Analytical Framework (method justification, full mechanism specification, observable implications, alternative mechanisms)
Findings I: Steps 1-2 of the mechanism
Findings II: Steps 3-4 of the mechanism and alternative explanations
Conclusion and Discussion
Bibliography

Example from the Literature

To see how process tracing works in practice, consider the treatment in George and Bennett (2005, ch. 10), which uses the end of the Cold War as an illustrative case. The question is how the Cold War ended peacefully — specifically, through what causal mechanism did Soviet foreign policy shift from confrontation to cooperation in the late 1980s? Rather than simply noting that Gorbachev came to power and the Cold War ended (a correlation), process tracing requires unpacking the intervening steps: the diffusion of new ideas among Soviet policy intellectuals, Gorbachev’s receptiveness to these ideas, specific policy decisions (such as accepting asymmetric arms reductions), and the feedback effects of Western responses. Each step is specified in advance and evaluated against the documentary and testimonial record. The analysis also considers alternative mechanisms — such as economic decline forcing Soviet concessions regardless of leadership — and shows why the evidence better supports an ideas-driven explanation.

This example illustrates several features that should appear in your own thesis. The mechanism is broken into discrete, testable steps rather than presented as a general narrative. Evidence is marshalled for each step individually, not lumped together. And the analysis explicitly engages with competing explanations rather than treating the preferred mechanism as the only possibility. For a concise overview of this approach, see Collier (2011), who uses simpler examples to demonstrate the same logic of diagnostic evidence and step-by-step evaluation.

Common Pitfalls

Confusing narrative with process tracing. Telling the story of what happened in chronological order is not process tracing. Process tracing requires a theorized mechanism specified in advance, explicit evidence tests, and systematic evaluation of each causal step. If your findings chapter reads like a history essay, you are narrating, not tracing.

Not specifying the mechanism in advance. If you derive your mechanism from the same evidence you use to test it, you are engaged in circular reasoning. The mechanism should come from theory (your literature review and analytical framework), not from your data. You can refine the mechanism during research, but the version you test must be stated before you present findings.

Cherry-picking evidence. Selecting only the evidence that supports your explanation while ignoring contradictory evidence undermines the entire exercise. Process tracing requires that you actively look for disconfirming evidence and report it transparently. If a step fails a hoop test, say so.

Ignoring alternative explanations. Showing that evidence is consistent with your mechanism is not enough. Evidence may also be consistent with other mechanisms. If you do not test alternatives, you cannot claim that your mechanism is the most plausible explanation. At minimum, identify the strongest competing explanation from the literature and subject it to the same evidence tests.

Treating all evidence as equal. A straw-in-the-wind test does not carry the same inferential weight as a smoking gun. Be explicit about what kind of test each piece of evidence constitutes. A thesis that relies entirely on straw-in-the-wind evidence is not compelling, even if there is a lot of it.

Vague or unfalsifiable mechanism steps. Each step must be specific enough that you could, in principle, find evidence against it. “Elites were unhappy” is too vague to test. “Business leaders in the export sector publicly opposed the sanctions policy” is testable.

Ask yourself

Read through your findings draft. For each paragraph, can you identify which step of the mechanism it addresses and what type of evidence test it represents? If not, you may be narrating rather than tracing.
Have you presented any evidence that contradicts your mechanism? If everything in your findings supports your argument perfectly, consider whether you have been sufficiently self-critical.

Key Readings

Essential

Beach, D., & Pedersen, R. B. (2019). Process-Tracing Methods: Foundations and Guidelines (2nd ed.). University of Michigan Press. — The most comprehensive methodological guide. Covers theory-testing, theory-building, and explaining-outcome variants in detail. Start here for method design.
Collier, D. (2011). Understanding process tracing. PS: Political Science & Politics, 44(4), 823–830. (Free PDF) — A clear, concise introduction to the method and the four evidence tests. Excellent starting point for students new to process tracing.

Advanced

Bennett, A., & Checkel, J. T. (Eds.). (2015). Process Tracing: From Metaphor to Analytic Tool. Cambridge University Press. — A multi-author volume that explores different traditions and applications of process tracing. Particularly useful for MA students seeking deeper methodological grounding.
Mahoney, J. (2012). The logic of process tracing tests in the social sciences. Sociological Methods & Research, 41(4), 570–597. — Formalizes the logic of the four evidence tests. Important for understanding the inferential foundations of the method.
George, A. L., & Bennett, A. (2005). Case Studies and Theory Development in the Social Sciences. MIT Press. — A foundational text on case study methods more broadly, with a landmark treatment of process tracing in the context of theory development. Chapter 10 is the key chapter.

Where to start: BA students should begin with Collier (2011) for a clear overview, then consult Beach and Pedersen (2019) for practical guidance on designing their analysis. MA students should read Beach and Pedersen in full and engage with Bennett and Checkel (2015) for a more advanced treatment.

Process tracing is often combined with other methods:

Comparative Case Study — Process tracing is a within-case method; comparative case study is a cross-case method. They answer different but complementary questions. Comparative design helps you identify which variables matter by examining patterns across cases. Process tracing helps you explain how those variables operate by tracing the causal mechanism within a single case. Many strong theses combine both.
Building a Corpus — Process tracing often requires assembling a structured body of primary source material — documents, speeches, policy records, media coverage. This guide covers practical strategies for collecting, organizing, and managing textual evidence systematically.