Counterfactual Data Research Articles

Simcausal R Package: Conducting Transparent and Reproducible Simulation Studies of Causal Effect Estimation with Complex Longitudinal Data.

The simcausal R package is a tool for specification and simulation of complex longitudinal data structures that are based on non-parametric structural equation models. The package aims to provide a flexible tool for simplifying the conduct of transparent and reproducible simulation studies, with a particular emphasis on the types of data and interventions frequently encountered in real-world causal inference problems, such as, observational data with time-dependent confounding, selection bias, and random monitoring processes. The package interface allows for concise expression of complex functional dependencies between a large number of nodes, where each node may represent a measurement at a specific time point. The package allows for specification and simulation of counterfactual data under various user-specified interventions (e.g., static, dynamic, deterministic, or stochastic). In particular, the interventions may represent exposures to treatment regimens, the occurrence or non-occurrence of right-censoring events, or of clinical monitoring events. Finally, the package enables the computation of a selected set of user-specified features of the distribution of the counterfactual data that represent common causal quantities of interest, such as, treatment-specific means, the average treatment effects and coefficients from working marginal structural models. The applicability of simcausal is demonstrated by replicating the results of two published simulation studies.

Calling in the Reserves: Judicial Replacements on the U.S. Supreme Court

The decision to recuse presents Supreme Court justices with a no-win choice between preserving legitimacy and risking the possibility of an evenly divided vote. A recent proposal would allow retired justices to serve when a justice is absent to help alleviate this tension. Here, we examine the potential implications of this proposal using more than 60 years of unique counterfactual data. Our results suggest that replacement justices could have changed the policy set by the Court in only a small subset of cases and, more importantly, could have helped the Court avoid more than a quarter of their even divisions.

Farewell's Linear Increments Model for Missing Data: The FLIM package

Missing data is common in longitudinal studies. We present a package for Farewell's Linear Increments Model for Missing Data (the FLIM package), which can be used to fit linear models for observed increments of longitudinal processes and impute missing data. The method is valid for data with regular observation patterns. The end result is a list of fitted models and a hypothetical complete dataset corresponding to the data we might have observed had individuals not been missing. The FLIM package may also be applied to longitudinal studies for causal analysis, by considering counterfactual data as missing data - for instance to compare the effect of different treatments when only data from observational studies are available. The aim of this article is to give an introduction to the FLIM package and to demonstrate how the package can be applied.

DO INDONESIAN MANAGERS TIME THEIR CORPORATE BONDS?

Corporate bonds are becoming popular in Indonesian capital market due to the ongoing decline in interest rates and the increase in credit rating. Debt Market Timing Theory argues that managers try to time their bonds issuance according to the market interest rate, relatively. This paper aims to analyze the debt issuance timing profile of Indonesian public listed companies. The samples are 24 bonds issuances, which have maturity period between 3 to 7, and companies already issued more than 1 bond issued within year 2009 and 2011. The manager’s behavior to time government bond rates is observed in the 5 working days window, whether the corporate bonds being issued at the lowest market interest rate (i.e., government bond rates) on the window. Bootstrap method is utilized to construct counterfactual data. The research finds that 7 out of 24 bonds issuance were issued at the lowest government bond yield within the window. Indonesian public listed companies had no ability to time their bond issuances during period 2009 until 2011. This paper reveals that the frequency of bond issuances made by each Indonesian company does not necessarily determine their capability to time government bond rates. However, bootstrap is a useful and more robust tool to help assessing the debt market timing ability when the samples taken are small in numbers.

Counter-factual scenario planning for long-range sustainable local-level tourism transformation

Many traditionally agricultural dependent economies have transformed towards service industries, such as tourism. This transformation resulted in significant impacts upon economies, communities and the environment. However, existing indicators to measure the impacts of tourism on regions have not been examined across the transformation process, leaving a gap in the understanding of long-range planning for tourism. Therefore, this paper investigates the relationship between economic, social and environmental indicators across the three main phases of tourism transformation. The three broad phases of tourism transformation can be observed to commence (inception), grow (construction) and then enter a steady state (urbanisation). To investigate this, our research surveyed 303 residents across three local areas at different stages of economic development and tourism dependency, to develop tourism, economic, social and environmental indicators. Employing counter-factual data, generalised ordinal logistic regression models were used to predict and compare the impact of changes in tourism on the economy, society and the environment across the regions. The findings indicate that the relationships between tourism and the triple bottom line differ depending on the stage of development. This research provides insight into local-level long-range planning, which can enable more sustainable tourism transformation, and explores avenues for future research.

Towards Detecting and Measuring Ballot Stuffing

This paper proposes a method for detecting electoral fraud in the form of ballot stuffing. As ballot stuffing increases both turnout and the incumbent‘s vote share in precincts where it occurs, precincts with low reported turnout are more likely to be clean. Information on clean precincts is used to simulate counterfactual data for “infected” precincts, which are then compared to the observed data. The method is applied to the 2006 Finnish presidential elections. The test fails to reject the hypothesis of no ballot stuffing for the original data, but detects artificially imputed 1.6% fraud. The same test implies that in the 2004 presidential elections in Russia at least 4.7% of the votes were stuffed in favor of the incumbent.

How do Agricultural Programmes Alter Crop Production? Evidence from Ecuador

Evaluating agricultural programmes requires considering not only the programmes’ influence on input and output indicators, but also considering the relationship between these indicators as embodied in the production technology. This article examines the impact on production of an intervention in the Ecuadorian Sierra designed to improve returns to potato production through training and through linking smallholders to high-value markets. Critical to identifying the impact of the programme is the careful construction of a counterfactual and meticulous data collection. To assess the impact of the programme on production, a weighted estimation, where weights are constructed through propensity score matching, is employed to estimate a production function within a damage abatement framework. The function incorporates a series of interaction terms to assess the impact of the programme on the production technology. The findings provide evidence that the programme enhances yields both through a general shift in technology as well as increased input use. The results suggest that the use of effective farming techniques that are learned through the programme induce this technological shift

Doubly Robust Estimation in Missing Data and Causal Inference Models

The goal of this article is to construct doubly robust (DR) estimators in ignorable missing data and causal inference models. In a missing data model, an estimator is DR if it remains consistent when either (but not necessarily both) a model for the missingness mechanism or a model for the distribution of the complete data is correctly specified. Because with observational data one can never be sure that either a missingness model or a complete data model is correct, perhaps the best that can be hoped for is to find a DR estimator. DR estimators, in contrast to standard likelihood-based or (nonaugmented) inverse probability-weighted estimators, give the analyst two chances, instead of only one, to make a valid inference. In a causal inference model, an estimator is DR if it remains consistent when either a model for the treatment assignment mechanism or a model for the distribution of the counterfactual data is correctly specified. Because with observational data one can never be sure that a model for the treatment assignment mechanism or a model for the counterfactual data is correct, inference based on DR estimators should improve upon previous approaches. Indeed, we present the results of simulation studies which demonstrate that the finite sample performance of DR estimators is as impressive as theory would predict. The proposed method is applied to a cardiovascular clinical trial.

Effectiveness of Potent Antiretroviral Therapy on Progression of Human Immunodeficiency Virus: Bayesian Modelling and Model Checking via Counterfactual Replicates

Summary We analyse data from a seroincident cohort of 457 homosexual men who were infected with the human immunodeficiency virus, followed within the multicentre Italian Seroconversion Study. These data include onset times to acquired immune deficiency syndrome (AIDS), longitudinal measurements of CD4+ T-cell counts taken on each subject during the AIDS-free period of observation and the period of administration of a highly active antiretro- viral therapy (HAART), for the subset of individuals who received it. The aim of the study is to assess the effect of HAART on the course of the disease. We analyse the data by a Bayesian model in which the sequence of longitudinal CD4+ cell count observations and the associated time to AIDS are jointly modelled at an individual subject’s level as depending on the treatment. We discuss the inferences obtained about the efficacy of HAART, as well as modelling and computation difficulties that were encountered in the analysis. These latter motivate a model criticism stage of the analysis, in which the model specification of CD4+ cell count progression and of the effect of treatment are checked. Our approach to model criticism is based on the notion of a counterfactual replicate data set Zc. This is a data set with the same shape and size as the observed data, which we might have observed by rerunning the study in exactly the same conditions as the actual study if the treated patients had not been treated at all. We draw samples of Zc from a null model M0, which assumes absence of treatment effect, conditioning on data collected in each subject before initiation of treatment. Model checking is performed by comparing the observed data with a set of samples of Zc drawn from M0.

Should the Treasury Price-Discriminate? A Procedure for Computing Hypothetical Bid Functions

Ever since FRIEDMAN [1959], it has been debated whether the Treasury should sell bonds in a uniform or in a discriminatory auction. Empirical research on this topic has been confined to experiments in which both auctions were used. But these experiments inherently contain an identification problem, since differences cannot be attributed to the auction format alone. We develop a method for generating counterfactual data on discriminatory auctions, using real data from uniform-price Treasury bond auctions in Switzerland. Our method allows us to investigate the performance of the two auctions without relying on experiments and without the identification problem. (JEL: D 44, H 63)