Fixed effects in rare events data
A Penalized Maximum Likelihood Solution
Bibliographic Data
| ID | 6342172 |
|---|---|
| Authors | Scott J Cook (0000-0003-0869-5137, Texas A&M University, corresponding author), Judith C Hays (0000-0003-2069-3038, University of Pittsburgh), Jude C Hays, Robert J Franzese (0000-0002-6614-4213, University of Michigan) |
| Year | 2020 |
| Volume | 8 |
| Issue | 1 |
| Pages | 92-105 |
| Publication date | 2020-01-01 |
| Peer Reviewed | Yes |
| Open Access | Yes |
| Type | ARTICLE |
| Venue | Political Science Research and Methods (JOURNAL) |
| Journal identifiers | ISSN: 2049-8470 • E-ISSN: 2049-8489 |
| Publisher | Cambridge University Press (CUP) (PUBLISHER) |
| DOI | 10.1017/psrm.2018.40 |
| OpenAlex | W2897354317 |
| Language | EN |
| Citations received | 39 |
| References cited | 31 |
Most agree that models of binary time-series-cross-sectional data in political science often possess unobserved unit-level heterogeneity. Despite this, there is no clear consensus on how best to account for these potential unit effects, with many of the issues confronted seemingly misunderstood. For example, one oft-discussed concern with rare events data is the elimination of no-event units from the sample when estimating fixed effects models. Many argue that this is a reason to eschew fixed effects in favor of pooled or random effects models. We revisit this issue and clarify that the main concern with fixed effects models of rare events data is not inaccurate or inefficient coefficient estimation, but instead biased marginal effects. In short, only evaluating event-experiencing units gives an inaccurate estimate of the baseline risk, yielding inaccurate (often inflated) estimates of predictor effects. As a solution, we propose a penalized maximum likelihood fixed effects (PML-FE) estimator, which retains the complete sample by providing finite estimates of the fixed effects for each unit. We explore the small sample performance of PML-FE versus common alternatives via Monte Carlo simulations, evaluating the accuracy of both parameter and effects estimates. Finally, we illustrate our method with a model of civil war onset
Binary number · Econometrics · Estimator · Event (particle physics · Fixed effects model · Monte Carlo method · Panel data · Random effects model · Rare events · Sample (material · Sample size determination · Statistics · Agricultural risk and resilience · Computer Science · International Development and Aid · Mathematics · Medicine · Political Conflict and Governance
Does External Environmental Uncertainty Encourage Policy Diffusion? Evidence from Chinese Administrative Approval Reform
The Futility of Shooting Down Strawmen
Democratic institutions and regulatory privileges for government debt
Inequality, Relative Deprivation, and Financial Distress
Global Markets and Local Representation
Economic development and revolutions. A cross-national investigation
Estimating group fixed effects in panel data with a binary dependent variable
Strike While the Iron is Hot
How contracted tree farmers engage in and benefit from inclusive value chains
Arming for conflict, arming for peace? How small arms imports affect intrastate conflict risk
Party Considerations for Renomination in Pre‐Electoral Coalitions
Who Sponsors the President's Program in the Opposition's House
Populism, Party Ideology, and Economic Expropriations
Politicians doing business
When and where we are
Electoral Politics, Fiscal Policy, and the Resource Curse
Debt Strain and Child Protective Services Involvement
Partial Effects for Binary Outcome Models with Unobserved Heterogeneity
Discriminatory coercion
Into the Fray
Multidimensional domestic gender inequality and the global diffusion of women’s ministries, 1975–2015
Do food safety certifications improve the safety of our food system? evidence from the U.S. Meat, Poultry, and egg industry
Not All Elections Are Created Equal
Locked Down, Lashing Out
Estimating Substantive Effects in Binary Outcome Panel Models
Countering Capture in Local Politics
Hot topics
Separation and Rare Events
Hypothesis Tests under Separation
Protest and state policy agendas
Armed Conflicts With Al-Qaeda and the Islamic State
The Law Behind Dispute Onset
Autocrats and Their Business Allies
Personalization of Power and Mass Uprisings in Dictatorships
Addressing Monotone Likelihood in Duration Modelling of Political Events
Central banks and civil war termination
Choosing tactics
The impact of son or daughter care on Chinese older adults' mental health
Nuclear balance and the initiation of nuclear crises
Bias reduction of maximum likelihood estimates
A weakly informative default prior distribution for logistic and other regression models
Consistent Estimates Based on Partially Consistent Observations
A solution to the problem of separation in logistic regression
The incidental parameter problem since 1948
Income and Democracy
The behaviour of the maximum likelihood estimator of limited dependent variable models in the presence of fixed effects
Analysis of Covariance with Qualitative Data
Greed and grievance in civil war
A warning on separation in multinomial logistic models
Economic Shocks and Civil War
Efficient Estimation of Time-Invariant and Rarely Changing Variables in Finite Sample Panel Analyses with Unit Fixed Effects
A Solution to Separation in Binary Response Models
Natural Disasters and the Risk of Violent Civil Conflict
Should I Use Fixed or Random Effects
Explaining Fixed Effects
Two Wrongs Make a Right
Dealing with Separation in Logistic Regression Models
Fixed-Effects Vector Decomposition
Sensitivity Analysis of Empirical Results on Civil War Onset
Do Ethnic and Nonethnic Civil Wars Have the Same Causes
Contagion or Confusion? Why Conflicts Cluster in Space
Proper Nouns and Methodological Propriety
Throwing Out the Baby with the Bath Water
Clear and Clean
Dirty Pool
How Foreign Aid Can Foster Democratization in Authoritarian Regimes
Ethnicity, Insurgency, and Civil War
| Unique citing works | 39 |
|---|---|
| Citations per year | 7,8 |
| Citation span | 2021 - 2026 (6) |
| Citation velocity | current |
| Highly cited | No |
| Citation types | Neutral: 39 |