Daina Chiba

Publications

  • 9. Military Coalitions and Crisis Duration [Show details]
        (with Jesse C. Johnson)
       Journal of Politics (forthcoming)
    • Abstract
      Forming a military coalition during an international crisis can improve a state's chances of achieving its political goals. We argue that the involvement of a coalition, however, can have unintended adverse effects on crisis outcomes by complicating the bargaining process and extending the duration of crises. This argument suggests that crises involving coalitions should be significantly longer than crises without coalitions. However, other factors that affect crisis duration are also likely to influence coalition formation. Therefore, taking into account the endogeneity of the presence of a coalition is essential to testing our hypothesis. To deal with this inferential challenge we develop a new statistical model that is an extension of instrumental variable estimation in survival analysis. Our analysis of 255 post-World War II interstate crises demonstrates that, even after accounting for the endogeneity of coalition formation, military coalitions tend to extend the duration of crises by approximately 284 days.
    • Publisher link
    • Replication materials
  • 8. Colonial Legacy and Foreign Aid: Decomposing the Colonial Bias [Show details]
        (with Tobias Heinrich)
       International Interactions (forthcoming)
    • Abstract
      It is well-known that donors give considerably more foreign aid to former colonies than to countries lacking past colonial ties. Unfortunately, we know relatively little about why this is the case. For one, there is almost never a theoretical justification for the inclusion of colonial history in statistical models. For the other, the only explicitly made rationale by Bueno de Mesquita and Smith (2009) actually predicts an interpretational problem: colonial history not only increases a former colony’s saliency to the donor, but also has left deep marks on recipients’ social and political institutions today. Both aspects shape how much aid a donor transfers to the recipient. This leaves ambiguous the meaning of the routinely found positive, sizable, and significant coefficient of colonial history on aid flows. We solve the inferential quandary by using a decomposition approach from labor econometrics. Our results show that about 75–100% of the colony effect on foreign aid stems from the greater saliency that donors give to policy concessions from former colonies.
    • Publisher link
    • Replication materials
  • 7. The Shape of Things to Come? [Show details]
        Expanding the Inequality and Grievance Model for Civil War Forecasts with Event Data
         (with Kristian Skrede Gleditsch)
       Journal of Peace Research, Vol. 54, No. 2 (Mar., 2017), pp. 275-297.
    • Abstract
      We examine if dynamic information from event data can improve on the efforts of Buhaug, Cederman and Gleditsch (2014) to predict civil war, using model with measures reflecting motivation and group characteristics relevant to civil war at the country level. The predictions from their model outperform conventional country-level models of civil war, emphasizing country characteristics, in an out-of-sample forecast. However, since most grievance measures change little over time, we surmise that a model reflecting potential motivation for conflict can be improved yet further with more dynamic information on mobilization and the behavior of actors. We find that a model with event data helps improve out-of-sample forecasts of civil war events in the Uppsala/PRIO Armed Conflict Data. Moreover, models with the original grievance measures do better than purely event based models, supporting our claim that both structure and event based components can add value to conflict prediction models.
    • Publisher link
    • Replication materials
  • 6. Careful Commitments: Democratic States and Alliance Design [Show details]
        (with Jesse C. Johnson and Brett Ashley Leeds)
       Journal of Politics, Vol. 77, No. 4 (Oct., 2015), pp. 968-982.
    • Abstract
      Evidence suggests that leaders of democratic states experience high costs from violating past commitments. We argue that because democratic leaders foresee the costs of violation, they are careful to design agreements they expect to have a high probability of fulfilling. This may cause democratic leaders to prefer flexible or limited commitments. We evaluate our argument by analyzing the design of alliance treaties signed by countries of the world between 1815 and 2003. We find that alliances formed among democratic states are more likely to include obligations for future consultation rather than precommitting leaders to active conflict, and defense pacts formed among democratic states are more likely to specify limits to the conditions under which member states must join their partners in conflict. This research suggests that separating screening effects and constraining effects of international agreements is even more difficult than previously believed. States with the greatest likelihood of being constrained are more carefully screened.
    • Manuscript in PDF
    • Publisher link
    • Web Appendix
    • Replication materials
  • 5. Every Story Has a Beginning, Middle, and an End (But Not Always in That Order): [Show details]
        Predicting Duration Dynamics in a Unified Framework
        (with Nils Metternich and Michael Ward)
       Political Science Research and Methods, Vol. 3, No. 3 (Sep., 2015), pp. 515-541.
    • Abstract
      There are three fundamental duration dynamics of civil conflicts: Time until conflict onset, conflict duration, and time until conflict recurrence. Theoretical and empirical models of war usually focus on one or at most two aspects of these three important duration dynamics. We argue that duration forecasting of conflict needs to incorporate all three conflict dynamics as unobserved factors might impact on all three of these dynamics. We present a new split-population seemingly-unrelated duration estimator that treats pre-conflict duration, conflict duration, and post-conflict duration as interdependent processes thus permitting improved predictions about the onset, duration, and recurrence of civil conflict. Our findings provide support for the more fundamental idea that prediction is dependent on a good approximation of the theoretically implied underlying data generating process. In addition, we account for the fact that some countries might never experience these duration dynamics or become immune after experiencing them in the past.
    • Manuscript in PDF
    • Publisher link
    • Replication materials
  • 4. A Copula Approach to the Problem of Selection Bias in Models of Government Survival [Show details]
        (with Lanny Martin and Randy Stevenson)
       Political Analysis, Vol. 23, No. 1 (Winter 2015), pp. 42-58.
    • Abstract
      Recent theories of coalition politics in parliamentary democracies suggest that government formation and survival are jointly determined outcomes. An important empirical implication of these theories is that the sample of observed governments analyzed in studies of government survival may be nonrandomly selected from the population of potential governments. This can lead to serious inferential problems. Unfortunately, current empirical models of government survival are unable to account for the possible biases arising from nonrandom selection. In this study, we use a copula-based framework to assess, and correct for, the dependence between the processes of government formation and survival. Our results suggest that existing studies of government survival, by ignoring the selection problem, significantly overstate the substantive importance of several covariates commonly included in empirical models.
    • Manuscript in PDF
    • Publisher link
    • Supplementary files
    • Replication materials
  • 3. Major Powers and Militarized Conflict [Show details]
       (with William Reed and Carla Martinez Machain)
       Journal of Conflict Resolution, Vol. 58, No. 6 (Sep., 2014), pp. 976-1002.
    • Abstract
      This paper attempts to answer the question of why major powers engage in more active foreign policy behaviors than minor powers. It does so by comparing two explanations for the increased conflict propensity of major powers. The first explanation focuses on major powers' observable capabilities, while the second stresses their different behavior. We incorporate both into an ultimatum model of conflict in which a state's cost of conflict consists of both observable and behavioral components. Using data from the period from 1870 to 2001, we empirically illustrate the observable and behavioral differences between major and minor powers. We then utilize a decomposition model to assess the relative significance of the two explanations. The results suggest that most of the difference in conflict propensity between major and minor powers can be attributed to observable differences.
    • Manuscript in PDF
    • Publisher link (journal subscription required)
    • Replication materials
  • 2. Institutional Opposition, Regime Accountability, and International Conflict [Show details]
       (with Songying Fang)
       Journal of Politics, Vol. 76, No. 3 (Jul., 2014), pp. 798-813.
    • Abstract
      Can international organizations constrain a leader's behavior during a military crisis? Existing studies have shown that joint membership in international organizations reduces the likelihood of dispute initiation; however, whether institutional opposition can prevent an ongoing conflict from escalating has yet to be investigated. We develop and test a theory of how domestic politics provides a mechanism through which international organizations can reverse the course of a military crisis. The argument leads to the hypothesis that more accountable regimes are less likely to escalate military crises when an international organization opposes their actions. We test the hypothesis with an analysis of territorial disputes from 1946 to 1995. We find that while neither institutional opposition nor the degree of regime accountability independently reduces the tendency for a country to escalate a conflict, the joint effect of the two does.
    • Manuscript in PDF
    • Publisher link (journal subscription required)
    • Online Appendix
    • Replication materials
  • 1. Decomposing the Relationship Between Contiguity and Militarized Conflict [Show details]
       (with William Reed)
       American Journal of Political Science, Vol. 54, No. 1 (Jan., 2010), pp. 61--73.  
    • Abstract
      It is well known that the majority of militarized conflicts and wars have been fought by neighbors. Yet, much remains to be learned about the relationship between shared borders and militarized conflict. This paper decomposes the effects of territorial contiguity into ex ante "observable" and "behavioral" effects. It provides powerful empirical evidence for the claim that although neighbors are more likely to experience conflict because of ex ante differences in observable variables such as economic interdependence, alliance membership, joint democracy, and the balance of military capabilities, most conflicts between neighbors occur because of differences in how neighbors and nonneighbors respond to the observable variables.
    • Manuscript in PDF
    • Publisher link (journal subscription required)
    • Replication materials

Resources

  • Some R files I prepared for teaching
  • LATEXresource for Rice students & scholars
    • I created some Beamer style files and presentation examples (slide and poster) for Rice students and scholars.
    • Link to the RiceBeamer page
  • Presenting statistical results
    • I ran a workshop for grad students and faculties in the political science department about how to present statistical results using R and Stata in combination. You can download this archived directory (zip, 45 MB) that contains the handout, an accompanying R code, Stata do-file, data, etc., necessary to replicate all the figures and tables I talked about in the workshop. Topics I covered include:
    • Producing publication-quality regression tables efficiently (Stata)
    • Calculating and plotting substantive effects (R) ··· Example 1, Example 2
    • Plotting descriptive statistics (R) ··· Example 3
    • Plotting regression coefficients (R) ··· Example 4
    • For more of these graphs, go to Using Graphs Instead of Tables website by Jonathan Kastellec and Eduardo Leoni.
  • R utilities for clustered standard error
    • As a quantitative IR researcher, I have seen many examples utilizing the following two estimation options of Stata -- , robust and , cluster(dyad). These frequently-used options are supposed to give us the "corrected" standard error estimates when observations are suspected to be interdependent within analysis unit (e.g., dyad) and/or distribution is not identical (i.e., errors are heteroskedastic). The good news is that the Stata implementation of these techniques is quite easy. The bad news is that this is a very handy tool for "star-gazing" -- an act of searching for statistical significance (stars!) by trying a bunch of different specifications and statistical patches without paying careful attention to the causal mechanism behind the data-generating process. Robust and/or clustered standard errors sometimes make your p-values much smaller, causing otherwise statistically insignificant coefficients to be very significant. Well, there is nothing inherently wrong about reporting robust standard errors instead of conventional ones. But, you should not use them just because this is the only way to have stars for your favorite variables. And you should not use them without understanding what they really mean.
    • So..., what exactly do we mean by standard error being "robust"? In what sense do we say that the estimation uncertainties are "adjusted for possible non-independence within dyad?" And, more (or less) importantly, how do we execute them on R? I suppose some step-by-step examples might help us gain intuition behind these methods, and will get us ready to implement the techniques on our own.
    • logit..., cluster(dyad) ··· Example R code to calculate clustered s.e. for a typical logit model of interstate disputes.
    • logit..., robust ··· Example R code to calculate robust s.e. for a typical logit model of conflict.
    • Sweave output (pdf) of the above two. This is generated by running the Sweave source (.Rnw) on R and then typesetting the tex output (.tex) on LATEX.
  • Riichi Mahjong Strategy Books


Daina


About

I am a senior lecturer (associate professor) in the Department of Government, University of Essex. Prior to joining Essex, I completed my graduate study in political science at Rice University and then worked as a postdoctoral associate at WardLab within the Department of Political Science, Duke University.


Conferences I'm Attending

News

  • July 2018: "Military Coalitions and Crisis Duration" is forthcoming at JoP.
  • Site Design