A magnifying glass focusing on an economic landscape, highlighting the uneven adoption of modern research tools.

Is the 'Credibility Revolution' in Economics Leaving Some Fields Behind?

"A new analysis reveals that while some areas of economics have embraced rigorous research methods, others are lagging, potentially impacting the reliability of their findings."


For the past two decades, economics has undergone a significant transformation known as the 'credibility revolution'. This movement emphasizes the use of transparent, credible research designs, leveraging new data to generate profound insights. Pioneered by figures like Joshua Angrist and Jörn-Steffen Pischke, the revolution aims to enhance the reliability and validity of economic research, addressing pressing questions from economic growth to the impacts of social and educational policies.

However, a recent study casts light on a concerning trend: the uneven adoption of these rigorous methods across different fields within economics. While some areas have fully embraced the credibility revolution, others are lagging behind, potentially undermining the robustness of their conclusions. This raises critical questions about whether the movement's initial momentum, identified by Angrist and Pischke, is still continuing apace, and whether certain empirical techniques are being favored over others.

Building on the work of Currie, Kleven, and Zwiers, this analysis examines the credibility revolution across various fields, including finance and macroeconomics. By analyzing over 32,000 National Bureau of Economic Research (NBER) working papers, the study identifies the frequency of phrases related to different empirical techniques, providing a comprehensive view of how these methods are being applied—or not—across the discipline.

AI Search Multiple angles on this topic

Measuring an Uneven Transformation

Quantifying the reach of the so-called 'credibility revolution' in economics is difficult, because no single authoritative measure tracks how widely its methods have been adopted across subfields. Available indicators, such as publication patterns, citation practices, or training changes, paint an incomplete and sometimes contradictory picture. Some observers suggest that fields closer to the design-based approaches favored by the movement have changed quickly, while others appear to have adapted more slowly. Given this uncertainty, any precise numbers offered about the movement's impact should be regarded as provisional rather than definitive.

Methods, Assumptions, and Their Boundaries

The accepted methods associated with the credibility revolution generally center on using natural experiments and other research designs to estimate causal effects, treating identification of a treatment effect as the core hurdle. In practice, these approaches require strong assumptions about comparability between treated and untreated groups, the validity of instruments, or the randomness of treatment assignment. Because those assumptions are rarely fully testable, results depend on judgment as much as on technique. A significant limitation is that the emphasis on credible identification can steer researchers toward questions that lend themselves to clean designs, even while many economically important questions do not.

The Roots of Credibility

Dictionaries consistently define credibility in terms of believability and trustworthiness, and these definitions are remarkably stable across reference works. Merriam-Webster, for instance, describes credibility as "the quality or power of inspiring belief," while Cambridge defines it as the fact that someone or something can be believed or trusted. Wikipedia extends the definition by identifying two key components, trustworthiness and expertise, each of which carries both objective and subjective dimensions. This shared foundation suggests that the concept at the heart of economics' credibility movement has long rested on a broadly agreed-upon meaning.

The Uneven Landscape of Credibility in Economics

A magnifying glass focusing on an economic landscape, highlighting the uneven adoption of modern research tools.

The study reveals that while the overall trends identified by Currie et al. continue to advance, significant heterogeneity exists across fields. Applied microeconomics has wholeheartedly embraced empirical techniques that emphasize research design, such as difference-in-differences, event studies, and randomized trials. In contrast, finance and macroeconomics are lagging in the uptake of these methods.

Within finance, corporate finance has shown robust growth in difference-in-differences designs, but the use of instrumental variables, regression discontinuity, and experimental methods remains limited. The adoption of popular tools like Bartik and shift-share instruments is also uneven, with rapid growth in applied micro areas like labor, trade, and economic history. Meanwhile, other tools, like synthetic controls, appear to have peaked in popularity.

  • Difference-in-differences dominate finance and macroeconomics.
  • Bartik and shift-share instruments are primarily used in applied micro areas.
  • Synthetic controls have plateaued in popularity.
AI Search Multiple angles on this topic

Emerging Work and Open Questions

Recent work in this space does not yet lend itself to a single summary, since research on the credibility movement is scattered across methodological reviews, replication studies, and field-specific appraisals. Much of the literature proceeds by example, documenting how particular natural experiments have reshaped understanding of specific economic questions. Reviews of the movement often note that its influence is measured more through practice than through formal pronouncements. Accordingly, general claims about the latest findings should be read as impressions of an ongoing conversation rather than settled conclusions.

When Trust Is Lost

At least one dictionary frames credibility in plainly behavioral terms, noting that a person "has credibility when you seem totally trustworthy or believable" and loses it by "lying, cheating and acting rather shady." This framing suggests that credibility is not fixed but earned and forfeited through conduct. In the context of economics, it implies that a research field's standing can erode when its practices, or the actions of its practitioners, stop inspiring belief. The definition offers a useful lens for understanding how failures, not just methodological advances, shape whether an approach is trusted.

Comparing Approaches Across Fields

Comparing the credibility movement across fields of economics is inherently difficult, because different areas face different data environments, institutional constraints, and traditions of evidence. Fields with abundant administrative or experimental data may have been easier for design-based methods to penetrate, whereas those reliant on structural modeling or historical narrative may pose greater challenges. Researchers offer only partial accounts of these differences, and rigorous cross-field comparisons are relatively scarce. Any strong claim that one field has been left behind should therefore be treated as a hypothesis in need of direct evidence.

These findings challenge the notion of a credibility revolution rapidly sweeping across all areas of economics. Instead, the picture is more nuanced, with the frontier of empirical work using credible, transparent research designs still centered in applied microeconomics, while other fields make progress at a slower pace. This raises concerns about the reliability of research in fields that have not fully embraced these rigorous methods.

The Path Forward: Diversifying Research Methods

The growing interest in and impact of difference-in-differences research across economics highlights how a single empirical technique, when widely adopted, can meaningfully shift the trajectory of an entire academic field. However, given some of the recent econometrics work flagging sensitivities and weaknesses in difference-in-differences, there may be value in researchers attempting to more broadly diversify their research methods portfolio. It is also quite striking that given the popularity of difference-in-difference that synthetic control methods have not grown further, as these methods have very similar properties.

AI Search Multiple angles on this topic

A Contested Legacy

Across various discussions of the credibility movement, a recurring theme is that its legacy is contested rather than settled. Supporters tend to emphasize gains in rigor and transparency, while critics point to what they see as narrowing or misdirected incentives. There appears to be little consensus among commentators about whether the movement has changed economics for the better overall. Given this range of views, expert commentary on the topic is best read as reflecting genuine disagreement rather than an established verdict.

What Comes Next

What the next phase of the credibility movement will bring remains genuinely uncertain. Some commentators anticipate continued refinement of identification strategies and greater emphasis on replication and transparency, while others foresee growing attention to areas where design-based methods are harder to apply. The frontier may also be shaped by new data sources, computational tools, and institutional incentives that are difficult to predict from current vantage points. Projections in this direction should therefore be framed as possibilities rather than forecasts.

Institutional Pressures and Structural Forces

Wider systemic forces, such as publication incentives, funding structures, and the training of new researchers, likely shape how the credibility movement unfolds in different fields, though these connections are not fully documented. When incentives reward papers with clean, novel identification, researchers may gravitate disproportionately toward such work. These pressures can interact with the methodological demands of particular fields in ways that are hard to isolate empirically. Any assessment of systemic challenges must acknowledge that much of this reasoning is necessarily speculative.

People, Practice, and Consequences

Behind the methodological debates are researchers making daily choices about research design, and the conclusions that emerge carry implications for real decisions in policy and business. How individuals weigh rigor against relevance, or career incentives against intellectual curiosity, is not well captured by survey statistics. The human costs of methodological drift, such as disaffected researchers or policy built on fragile findings, are largely anecdotal at this stage. These human dimensions warrant attention even though they resist straightforward measurement.

About this Article -

Written with AI assistance from published research, and reviewed by the Mystum team. See our About page for more information.

This article is based on research published under:

DOI-LINK: https://doi.org/10.48550/arXiv.2405.20604,

Title: Tracking The Credibility Revolution Across Fields

Subject: econ.gn q-fin.ec

Authors: Paul Goldsmith-Pinkham

Published: 30-05-2024

Everything You Need To Know

1

What is the 'credibility revolution' in economics and why is it important?

The 'credibility revolution' is a significant transformation in economics, emphasizing transparent and credible research designs. This movement, championed by figures like Joshua Angrist and Jörn-Steffen Pischke, focuses on enhancing the reliability and validity of economic research. It leverages new data and rigorous methods to provide profound insights into various economic issues, such as economic growth and the impacts of social and educational policies. The importance lies in ensuring that economic research findings are trustworthy and can inform effective policy decisions. Without embracing credible methods, the conclusions drawn might be unreliable, leading to flawed understanding and potentially detrimental policy implementations.

2

How does the adoption of the 'credibility revolution' vary across different fields within economics?

The adoption of rigorous research methods from the 'credibility revolution' varies significantly across different fields. Applied microeconomics has widely embraced empirical techniques like difference-in-differences, event studies, and randomized trials. However, finance and macroeconomics have been slower to adopt these methods. While corporate finance shows growth in difference-in-differences designs, other methods like instrumental variables, regression discontinuity, and experimental methods are less common. This uneven adoption indicates that the credibility revolution's influence isn't uniform across all areas of economics, impacting the reliability of research outcomes differently depending on the field.

3

What are 'difference-in-differences' and 'synthetic control' methods, and how are they used in economic research?

'Difference-in-differences' is an empirical technique used to estimate the causal effect of a treatment by comparing the changes over time in an outcome variable for a group that is exposed to the treatment (the treatment group) to the changes over time in the same outcome variable for a group that is not exposed to the treatment (the control group). 'Synthetic control' is a method used to estimate the effect of an intervention (such as a policy change) by comparing the outcomes in a treated unit to a synthetic control group. This synthetic control group is constructed as a weighted average of control units, designed to closely resemble the treated unit before the intervention occurred. Both methods are examples of the types of research design that the 'credibility revolution' emphasizes. The choice and implementation of these methods are crucial for ensuring the validity and reliability of research findings.

4

Why is it concerning that finance and macroeconomics are lagging in adopting rigorous research methods?

It is concerning because the slow adoption of rigorous research methods in finance and macroeconomics could undermine the reliability of research findings in these fields. Without embracing the transparent and credible research designs promoted by the 'credibility revolution', conclusions may be less robust and potentially misleading. This could have significant implications for policy decisions and economic understanding. For example, if research in macroeconomics relies on less rigorous methods, policymakers might make decisions based on flawed evidence, leading to ineffective or even harmful economic policies. Similarly, in finance, unreliable research can lead to poor investment strategies and market instability. Therefore, the uneven adoption of rigorous methods across economics raises questions about the credibility of research and its implications for practical applications and broader societal impact.

5

What are 'Bartik' and 'shift-share instruments,' and in which areas of economics are they primarily used?

'Bartik' and 'shift-share instruments' are both empirical techniques used in economic research to estimate the causal effects of various economic interventions or events. Bartik instruments, also known as 'supply-side instruments,' often use initial conditions and national trends to predict economic changes in local areas. Shift-share instruments are often used to predict local economic outcomes by combining national trends with a local area's industry composition. They are primarily used in applied micro areas such as labor economics, trade, and economic history. Their adoption is uneven across different fields, with rapid growth in applied micro areas. Understanding the applications of these instruments highlights the varied approaches within the 'credibility revolution' and how certain fields are more actively embracing them than others.

Newsletter Subscribe

Subscribe to get the latest articles and insights directly in your inbox.