Assessing causal inference problems with Bayesian process tracing: the economic effects of proportional representation and the problem of endogeneity

ABSTRACT

In recent years, causal process tracing is becoming increasingly formalised as scholars have begun to follow Bayesian logic and thereby manage to combine the interpretative and contextual nuance of older forms of process tracing with the inferential rigours of Bayesian analysis. This paper illustrates the basic logic of Bayesian process tracing by drawing on the political economy literature that studies the social policy and economic effects of electoral systems. It compares and contrasts how each approach deals with the problem of endogeneity.     

Extrapolation of causal effects – hopes,assumptions, and the extrapolator’s circle

ABSTRACT

I consider recent strategies proposed by econometricians for extrapolating causal effects from experimental to target populations. I argue that these strategies fall prey to the extrapolator’s circle: they require so much knowledge about the target population that the causal effects to be extrapolated can be identified from information about the target alone. I then consider comparative process tracing (CPT) as a potential remedy. Although specifically designed to evade the extrapolator’s circle, I argue that CPT is unlikely to facilitate extrapolation in typical econometrics and evidence-based policy applications. To argue this, I offer a distinction between two kinds of extrapolation, attributive and predictive, the latter being prevalent in econometrics and evidence-based policy. I argue that CPT is not helpful for predictive extrapolation when using the kinds of evidence that econometricians and evidence-based policy researchers prefer. I suggest that econometricians may need to consider qualitative evidence to overcome this problem.     

Process tracing of extensive and intensive processes

ABSTRACT

This essay briefly analyzes the introduction of process tracing to the political science and political economy methodological toolkit. It then proposes a theory-guided variant of process tracing and distinguishes between its application in extensive and intensive type of processes. It argues that the comparative advantage of process tracing vis-à-vis other social sciences research methods lies in its potential to uncover the causal mechanisms that link the constitutive events of intensive type of processes. It shows that theory-guided process tracing of intensive processes can be successfully applied to illustrate, test and produce theory.     

Between X and Y: how process tracing contributes to opening the black box of causality

ABSTRACT

This article maps the methodological debate on process tracing and discusses the diverse variants of process tracing in order to highlight the commonalities beyond diversity and disagreements. Today most authors agree that process tracing is aimed at unpacking causal and temporal mechanisms. The article distinguishes two main types of use for process tracing. Some are more inductive, aimed at theory building (i.e. at uncovering and specifying causal mechanisms) while others are more deductive, aimed at theory testing (and refining). The paper summarizes the main added value and drawbacks of process tracing. It ends by providing ten guidelines for when and how to apply process tracing.     

It's all about mechanisms – what process-tracing case studies should be tracing

ABSTRACT

Process-tracing (PT) as a distinct case-study methodology involves tracing causal mechanisms that link causes (X) with their effects (i.e. outcomes) (Y). We trace causal mechanisms whereby a cause (or set of causes) produces an outcome to both: (1) make stronger evidence-based inferences about causal relationships because the analysis produces within-case evidence of each step of the causal process (or absence thereof) in between a cause and outcome, and (2) because tracing mechanisms gives us a better understanding of how a cause produces an outcome. Yet, when we look at the methodological literature on PT, there is considerable ambiguity and discord about what causal mechanisms actually are. The result of this ambiguity and discord about what mechanisms are clearly maps onto existing applications of PT, with most PT case studies completely ignoring the underlying theoretical causal processes. In the few PT applications where mechanisms are unpacked, they are typically only developed in a very cursory fashion, with the result that there is considerable ambiguity about what theoretical process the ensuing case study actually is tracing. If we want to claim we are tracing causal mechanisms, the causal processes in between X and Y need to be unpacked theoretically. How can we claim we are tracing a causal ’process’ when we are not told what the process (i.e. mechanism) actually is? To alleviate this problem, the article attempts to develop a clearer definition of causal mechanisms to provide scholars with a framework for theorising mechanisms in a fashion that is amenable to in-depth empirical analysis using PT.     

Narrating Crisis,Constructing Policy: Economic Ideas and Institutional Change in Syria

During crises, ideas play a decisive role in shaping radical paradigm shifts in economic governance. However, not all crises immediately produce such ‘great transformations’. Why do some ideas result in incremental rather than abrupt change after crisis? To identify mechanisms potentially explaining this variation, I conduct an exploratory process tracing of an understudied case of incremental institutional change: post-independence Syria. Competing political actors in Syria converged on identical policy responses to crisis despite their very different interpretations of its causes. Although power oscillated between these increasingly bitter rivals in the early 1950s, their ideational consensus on economic issues nevertheless led to a decade of steady institutional change that transformed previously fragile government institutions into powerful vehicles of statism. I derive from this analysis the potential causal significance of two new variables – crisis narrative and crisis response – and hypothesise that their configuration can explain variation in post-crisis patterns of institutional change. Ideas can explain not only the new direction of economic governance after crisis, but also the speed and scale of its movement.     

Process tracing,abstraction, and varieties of cognitive interest

ABSTRACT

This comment responds specifically to some points made by Trampusch and Palier in this volume, particularly the distinction between inductive and deductive approaches, and the inquiry into the ‘added value’ of process tracing.     

Why Companies Bring the State Back In. The Voluntary Self-Nationalisation of the Anglo-Persian Oil Company and the Rise of ‘Governance by Government’

ABSTRACT

While companies usually prefer private to hybrid governance, they sometimes transfer governance to governments. This paper assumes that this emerges from a decline in a firm’s relative market position. It tests this assumption in a two-stage qualitative analysis: First, it conducts a fuzzy-set QCA of these decisions based on BP’s Annual Company Reports. Second, it presents a process tracing of the voluntary self-nationalisation of the Anglo-Persian Oil Corporation. The paper produces three core findings: First, threats to a company’s survival enable a preference reversal in favour of governmentalisation. Second, shareholder dissatisfaction is a crucial motivator for governmentalisation. Third, managers with an entrepreneurial role-model are more sensitive to autonomy costs and more likely to opt for unconventional governance options.     

Oil shocks and stock volatility: new evidence via a Bayesian,graph-based VAR approach

ABSTRACT

This article examines the temporal dependence between three oil shocks and realized volatility in the stock markets of G20 countries between 1994 and 2019. By applying a novel, graphical, Bayesian VAR (BGVAR) model, we calculate unidirectional linkages of oil and stock volatility with a full and segmented sample. The results suggest an overall causality from stock volatility to oil shocks. For certain short, specific periods, the causal direction reverses. Depending on the country and the source of an oil shock, the magnitude and type of the effect can vary considerably. Specific oil-market shocks occur most often in our full sample. In a time-varying structure, oil supply shocks’ impact on stock volatility is more prominent, and net oil-importing countries’ responses to these shocks are greater than for oil-exporting countries. In addition, we find that relationship dynamics can capture market information, such as global economic growth during the 2008–2009 financial crisis.     

A qualitative cross-impact approach to find the key technology

This article presents a qualitative analysis method based on fuzzy relations for a cross-impact model designed for a technology impact assessment. The cross-impact knowledge is often uncertain or fuzzy when dealing with future events. Assessing the cross-impact relationships among future technologies creates a more uncertain or fuzzy situation because of the time and the uncertainty involved in evaluating future technologies. In addition, experts prefer to use linguistic terms or fuzzy values in their predictions. Thus a cross-impact matrix is represented as fuzzy relations on causal concepts. We therefore develop inference algorithms based on fuzzy relations and show a simple technology assessment example to illustrate this approach. This approach is useful in finding the key technology because it considers not only the direct impact but also the indirect impact.     

Bayesian inference in cointegrated systems

In this paper I deal with Bayesian methods for conducting inference on important features of (potentially) cointegrated VAR models involving I(1) variables. Firstly, (informal) inference is made on the cointegrating rank of the system. Secondly, posterior analysis is used to verify the validity of over-identifying restrictions on the cointegration parameters. Thirdly, posterior distributions are obtained for impulse response functions and predictive densities at different horizons. The relevant posterior distributions are obtained by means of Monte Carlo integration. The analysis is based on the use of simple weakly informative priors. Two applications on simulated data and on the Danish money demand data are presented.     

Testing fiscal sustainability in Poland: a Bayesian analysis of cointegration

Fiscal sustainability is a central topic for most of the transition economies of Eastern Europe. This paper focuses on a particular country: Poland. The main purpose is to investigate, empirically, whether the post-transition fiscal policy is consistent with the intertemporal budget constraint, used as a formal theoretical framework. To test debt stabilization, the empirical analysis is made in two steps: first, we use a Bayesian methodology to conduct inference about the cointegrating relationship between budget revenues and (inclusive of interest) expenditures and to select the cointegrating rank. Second, we apply Bayesian inference to the estimation of the cointegrating vector and of the adjustment parameters. With a single cointegrating relation, we make use of some known results concerning the posterior density of the cointegrating vector, which belongs to the poly-t densities class. In this way, we experiment the usefulness of Bayesian inference in precisely assessing the magnitude of the cointegrating vector. Moreover, we show to what extent the likelihood of the data is important in revising the available prior information, relying on numerical integration techniques.     

Economic policy uncertainty,CEO characteristics and firm R&D expenditure: a Bayesian analysis

ABSTRACT

This paper applies a linear Bayesian regression model to study the effects of economic policy uncertainty (EPU) and Chief Executive Officer (CEO) characteristics on firm Research and Development (R&D) expenditure. We specifically analysed data from 1,163 Chinese companies listed as A-shares from 2008–2016. EPU is believed to curtail firm investments as it causes unexpected market conditions. Yet, our findings obtained with Bayesian analysis show a positive relationship between EPU and firm R&D expenditure. Specifically, we found that some CEO attributes (e.g., age, tenure and marketing/sales experience) led to a negative relationship between EPU and firm R&D expenditure, whereas other attributes (e.g., education, overseas study/work experience, product R&D experience, and process engineering experience) lead to a positive relationship. Our findings provide nuanced insights into how different CEO characteristics influence firms’ R&D expenditure in a context of uncertainty.     

On the uniqueness of Groves mechanisms and the payoff equivalence principle

Consider a standard mechanism design setting with quasi-linear preferences and private valuations. From Holmström (1979), we know that if the valuations are smooth with respect to types then any efficient, dominant strategy mechanism is in the class of Groves mechanisms. Here I show that, given regular assumptions on the primitives of the design problem, a weaker condition that includes the case of non-smooth valuations is sufficient and necessary for the uniqueness of Groves mechanisms among all efficient, dominant strategy mechanisms. This condition, which imposes a restriction on the behavior of the one-sided directional derivatives of the valuation functions with respect to individual types, is also shown to be sufficient and necessary to obtain the Payoff Equivalence principle for dominant strategy mechanisms whose choice rules are affine maximizers.     

Clinical trials with multiple endpoints can establish a correlation,but not (yet) causality,between dietary supplementation with omega-3 fatty acids and keratoconjunctivitis sicca

Aim: We review clinical evidence of therapeutic efficacy and effectiveness of omega-3 fatty acids (omega-3s) in keratoconjunctivitis sicca, colloquially known as dry eye disease. In doing so, we identify relevant literature to address the following questions: (1) What definitive guidance can clinical evidence offer eye physicians and their patients? (2) What aspects of omega-3 supplementation lack definitive evidence, and how might economic assessments help?

Methods: A targeted and systematic search strategy based on PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) was designed in this study for refereed clinical trials of omega-3s in dry eye treatment. Four key databases were accessed. Records were filtered through a three-step process using predetermined inclusion criteria. Data was extracted for experimental design, sample population characteristics, content of test and control groups, symptoms and associated opthalmologic conditions, diagnostic measures, patient health outcomes, adverse events, and model time horizon.

Results: A total of 219 records were initially identified, of which 21 prospective clinical trials, with a total of 2,973 participants, were admitted for review. Clinical evidence indicates that daily oral supplementation with omega-3s statistically correlates with dry eye treatment in the general symptomatic population as well as induced sub-populations. Reported adverse events are minimal. These trials nonetheless exhibit considerable heterogeneity in clinical endpoints, particularly as a result of the multi-factorial character of dry eye as well as continuous advances in scientific knowledge and technology. Their findings and recommendations appear to be of limited external validity. And causal inferences are needed, but difficult to establish. These have encouraged and sustained wide variations in ophthalmologic practice and normative decision-making.

Conclusions: Comparability of omega-3 therapeutic efficacy and effectiveness remains a major challenge in dry eye disease management. Notwithstanding its multi-factorial character, addressing prevailing methodological and empirical issues in clinical trials will help reduce knowledge gaps and normative choices among eye physicians and patients alike. In this regard, pharmacoeconomics offers a useful and robust toolset through which analysis of cost-minimization, cost-effectiveness, and cost-utility as well as meta-analysis can be comparatively pursued. Blending measures of costs and outcomes puts in perspective the heterogeneity of clinical endpoints in keratoconjunctivitis sicca.     

Identifying Neo-Schumpeterian Innovation in Service Firms: A Conceptual Essay with a Novel Classification

The identification of innovation in service firms is problematic since there is no consensus of opinion on its conceptualization. Recent papers suggest both distinctive features of innovation in services and distinctive types of service innovation. This article reviews and evaluates these findings from a Schumpeterian perspective. The evaluation justifies conceptualizing service innovation as a specific case of service development with a reference to Schumpeter, but not as strict as proposed by Drejer (2004) [Drejer, I. (2004) Identifying Innovation in Surveys of Services: A Schumpeterian Perspective. Research Policy, 33, 551–562]. Despite the simultaneity of production and consumption in services, this article claims that the distinction between product innovation and process innovation should be preferred to other ways of classifying innovation in service firms. Finally, changes in the denomination of services are advanced as a key to the identification of development and innovation in service firms.     

How historical analogies in newspapers of five countries make sense of major events: 9/11, Mumbai and Tahrir Square

We analyze how historical analogies are used in the media to make sense of novel events. While earlier work focused on single case studies, this is the first quantitative analysis comparing historical analogies invoked in three events in newspapers from five countries. With very high intercoder reliability we found 881 invocations of historical analogies. We found an interesting contrast between the roles of historical analogies in foreign policy decision making vs. newspaper articles. When the task is advocacy for policy choice, a compelling historical analogy will be one in which the causal mechanisms are as similar as possible to the current situation so that similar actions are likely to lead to similar results. Instead, newspapers spend more time at the early stages of sense-making and help the audience understand just a few features of the current situation. Newspapers thus offer a much broader range of historical analogies without much regard to maximizing similarity.     

Reworking wild bootstrap-based inference for clustered errors

Cluster-robust inference is increasingly common in empirical research. With few clusters, inference is often conducted using the wild cluster bootstrap. With conventional bootstrap weights the set of valid $P$-values can create ambiguities in inference. I consider several modifications to the bootstrap procedure to resolve these ambiguities. Monte Carlo simulations provide evidence that both a new 6-point bootstrap weight distribution and a kernel density estimation approach improve the reliability of inference. A brief empirical example highlights the implications of these findings.     

The evolutionary approach to entropy: Reconciling Georgescu-Roegen's natural philosophy with the maximum entropy framework

The paper explores the relevance of recent developments in the Maximum Entropy hypothesis for reinstating Georgescu-Roegen's natural philosophy, with special emphasis on the concepts of evolution and time. The key point is the naturalization of the notion of ‘subjectivity’ in both the Georgescu-Roegen framework and Jaynes's subjectivistic interpretation of thermodynamics and statistical mechanics. I introduce the concept of ‘observer relativity’ with reference to the evolution of ‘physical inference devices’. Then, the MaxEnt formalism can be understood as a principle underlying natural selection. Further, given natural selection, maximum entropy production (MEP) results from the confluence of maximum power (Lotka) and the maximization of information capacity, driven by energy dispersal. In these processes, hierarchical structures of gradients of energy dissipation reflect alternative positions of system boundaries, and hence different perspectives of observer-relativity. Thus, I can distinguish between observer relative Entropy_OR and observer independent Entropy_OI. This allows to reconstruct conceptually the two notions of time proposed by Georgescu-Roegen, with subjectivistic time seen as time relative to the evolutionary process involving incommensurable qualitative change. I claim that this philosophical view offers a powerful conceptual framework for recent empirical research into the energetics of economic growth.