# Bayesian Spatial Split-Population Survival Model with Applications to Democratic Regime Failure and Civil War Recurrence

^{1}

^{2}

^{*}

## Abstract

**:**

## 1. Introduction

## 2. (Spatial) Split-Population Survival Model

#### 2.1. Model Development

#### 2.2. Markov Chain Monte Carlo Estimation

- Choose a starting point ${\beta}_{0},{\gamma}_{0}$, ${\rho}_{0}$, ${\lambda}_{0}$ and corresponding ${\mathbf{W}}_{0}=\{{W}_{1},\dots ,{W}_{N}\}$ and ${\mathbf{V}}_{0}=\{{V}_{1},\dots ,{V}_{N}\}$, then set $k=0$.
- Update ${\mathsf{\Sigma}}_{\beta}\sim \pi \left({\mathsf{\Sigma}}_{\beta}\right|\beta )$, ${\mathsf{\Sigma}}_{\gamma}\sim \pi \left({\mathsf{\Sigma}}_{\gamma}\right|\gamma )$, $\lambda \sim \pi \left(\lambda \right|\mathbf{W},\mathbf{V})$ using Gibbs sampling. The closed form of the full conditional distributions for $\pi \left({\mathsf{\Sigma}}_{\beta}\right|\beta )$, $\pi \left({\mathsf{\Sigma}}_{\gamma}\right|\gamma )$, $\pi \left(\lambda \right|\mathbf{W},\mathbf{V})$ are derived and defined in the Supplementary Materials.
- Update $\beta \sim \pi \left(\beta \right|\mathbf{C},\mathbf{X},\mathbf{Z},\mathbf{t},\mathbf{W},\mathbf{V},\gamma ,\rho ,{{}^{\u2013}}_{\beta},{\mathsf{\Sigma}}_{\beta})$, $\gamma \sim \pi \left(\gamma \right|\mathbf{C},\mathbf{X},\mathbf{Z},\mathbf{t},\mathbf{W},\mathbf{V},\beta ,\rho ,{{}^{\u2013}}_{\gamma},{\mathsf{\Sigma}}_{\gamma})$, and $\rho \sim \pi \left(\rho \right|\mathbf{C},\mathbf{X},\mathbf{Z},\mathbf{t},\mathbf{W},\mathbf{V},\beta ,\gamma ,{a}_{\rho},{b}_{\rho})$ using the slice sampler with stepout and shrinkage (Neal, 2003); see the Supplementary Materials for details on performing the slice sampling operation in this step.
- Update $\mathbf{W}\sim \pi \left(\mathbf{W}\right|\mathbf{C},\mathbf{X},\mathbf{Z},\mathbf{t},\mathbf{V},\beta ,\gamma ,\rho ,\lambda )$ and $\mathbf{V}\sim \pi \left(\mathbf{V}\right|\mathbf{C},\mathbf{X},\mathbf{Z},\mathbf{t},\mathbf{W},\beta ,\gamma ,\rho ,\lambda )$ via Metropolis–Hastings.
- Set $k=k+1$, then return to Step 2 and repeat for K iterations.

- Choose the initial values of $\beta ,\gamma $, and $\rho $, then set $m=0$.
- Update ${\mathsf{\Sigma}}_{\beta}$ and ${\mathsf{\Sigma}}_{\gamma}$ via Metropolis–Hastings; see the Supplementary Material for the closed form of the full conditional distributions for ${\mathsf{\Sigma}}_{\beta}$ and ${\mathsf{\Sigma}}_{\gamma}$.
- Update $\beta $, $\gamma $, and $\rho $ using the slice sampler with stepout and shrinkage, as described in the Supplementary Materials.
- Repeat Steps 2 and 3 until the chain converges.
- After M iterations, summarize the parameter estimates using posterior samples.

## 3. Monte Carlo Simulations

## 4. Empirical Applications

#### 4.1. Democratic Consolidation and Survival

**A**. We construct a matrix

**A**with elements ${a}_{i{i}^{\prime}}$ such that ${a}_{i{i}^{\prime}}=1$ for each year if the capital of country i is less than 800 km from the capital of country ${i}^{\prime}$ and ${a}_{i{i}^{\prime}}=0$ if countries i and ${i}^{\prime}$ are greater than 800 km from each other. Using geographic proximity as the spatial relationship of interest is appropriate, as it allows the frailties to be correlated with those of neighboring democracies rather than assuming spatial independence even within the same regions. Considering our Bayesian MCMC estimation approach, we incorporate the spatial information in

**A**by employing separate CAR priors for the frailty terms vector

**V**(split-stage) and

**W**(survival-stage), which implies a CAR structure of $\mathbf{V}|\lambda $∼ CAR$\left(\lambda \right)$ and $\mathbf{W}|\lambda $∼ CAR$\left(\lambda \right)$. The spatial SP Weibull model is estimated based on the sample from [17] using the MVN prior and our MCMC algorithm described earlier and assigning the Gamma hyperprior for $\lambda $. Here, we use the hyperparameters $a=1$, $b=1$, ${S}_{\beta}$ = ${I}_{p1}$, ${S}_{\gamma}$ = ${I}_{p2}$, ${\nu}_{\beta}=p1$, ${\nu}_{\gamma}=p2$. Recall that ${\mathsf{\Sigma}}_{\beta}$ is the variance of the MVN prior of the vector $\beta $ for p1-dimensional survival stage covariates and that ${\mathsf{\Sigma}}_{\gamma}$ is the MVN’s prior of the vector $\gamma $ for p2-dimensional split-stage covariates. Hence, when we employ the Inverse Wishart (IW) distribution to estimate both ${\mathsf{\Sigma}}_{\beta}$, in which ${\nu}_{\beta}$ is the hyperparameter, and ${\mathsf{\Sigma}}_{\gamma}$, in which ${\nu}_{\gamma}$ is the hyperparameter, we adopt the values p1 for ${\nu}_{\beta}$ and p2 for ${\nu}_{\gamma}$. Finally, $\lambda \sim \mathrm{Gamma}({a}_{\lambda},{b}_{\lambda})$ with a vague prior $({a}_{\lambda},{b}_{\lambda})=(0.001,1/0.001)$. Our Bayesian SP Weibull model results are based on a set of 50,000 iterations after 4000 burn-in scans and thinning of 10.

#### 4.2. Post-Civil War Peace Duration

**V**) range from −1.19 to 1.49 with a standard deviation of 0.5503, and the survival-stage frailties (

**W**) range from −1.21 to 0.727 with a standard deviation of 0.4498. In both stages, there seem to be regional clusters of frailty values.

## 5. Discussion and Conclusions

## Supplementary Materials

## Author Contributions

## Funding

## Data Availability Statement

## Acknowledgments

## Conflicts of Interest

## Abbreviations

CAR | Conditionally Autoregressive |

CP | Convergence Probabilities |

d.g.p. | Data Generation Process |

i.i.d. | Independent and Identically Distributed |

IW | Inverse Wishart |

MC | Monte Carlo |

MCMC | Markov Chain Monte Carlo |

MCSE | Monte Carlo Standard Error |

MVN | Multivariate Normal |

NS | Non-Spatial |

NSF | Non-Spatial Frailty |

RMSE | Root Mean Square Error |

SP | Split Population |

## Appendix A

**Figure A1.**Results from the two autocorrelation diagnostics for the democratic survival application: (

**a**) join count and (

**b**) Moran’s I.

**Figure A2.**Results from the two autocorrelation diagnostics for the post-war peace duration application: (

**a**) join count and (

**b**) Moran’s I.

## References

- Box-Steffensmeier, J.M.; Jones, B.S. Event History Modeling: A Guide for Social Scientists; Cambridge University Press: New York City, NY, USA, 2004. [Google Scholar]
- Wang, Y.; Klijn, J.G.; Zhang, Y.; Sieuwerts, A.M.; Look, M.P.; Yang, F.; Talantov, D.; Timmermans, M.; Meijer-van Gelder, M.E.; Yu, J.; et al. Gene-expression profiles to predict distant metastasis of lymph-node-negative primary breast cancer. Lancet
**2005**, 365, 671–679. [Google Scholar] [CrossRef] [PubMed] - Bremhorst, V.; Lambert, P. Flexible estimation in cure survival models using Bayesian P-splines. Comput. Stat. Data Anal.
**2016**, 93, 270–284. [Google Scholar] [CrossRef] - Clark, D.H.; Regan, P.M. Opportunities to fight: A statistical technique for modeling unobservable phenomena. J. Confl. Resolut.
**2003**, 47, 94–115. [Google Scholar] [CrossRef] - Ray, R.A.; Perry, R.W.; Som, N.A.; Bartholomew, J.L. Using cure models for analyzing the influence of pathogens on salmon survival. Trans. Am. Fish. Soc.
**2014**, 143, 387–398. [Google Scholar] [CrossRef] - Schmidt, P.; Witte, A.D. Predicting criminal recidivism using ‘split population’survival time models. J. Econom.
**1989**, 40, 141–159. [Google Scholar] [CrossRef] - Peng, Y.; Taylor, J.M. Mixture cure model with random effects for the analysis of a multi-center tonsil cancer study. Stat. Med.
**2011**, 30, 211–223. [Google Scholar] [CrossRef] - Patilea, V.; Van Keilegom, I. A general approach for cure models in survival analysis. Ann. Stat.
**2020**, 48, 2323–2346. [Google Scholar] [CrossRef] - Bagozzi, B.E.; Joo, M.M.; Kim, B.; Mukherjee, B. A Bayesian Split Population Survival Model for Duration Data With Misclassified Failure Events. Political Anal.
**2019**, 27, 415–434. [Google Scholar] [CrossRef] - Chiba, D.; Metternich, N.W.; Ward, M.D. Every story has a beginning, middle, and an end (but not always in that order): Predicting duration dynamics in a unified framework. Political Sci. Res. Methods
**2015**, 3, 515–541. [Google Scholar] [CrossRef] - Banerjee, S.; Carlin, B.P. Parametric spatial cure rate models for interval-censored time-to-relapse data. Biometrics
**2004**, 60, 268–275. [Google Scholar] [CrossRef] - Darmofal, D. Bayesian Spatial Survival Models for Political Event Processes. Am. J. Political Sci.
**2009**, 53, 241–257. [Google Scholar] [CrossRef] - Darmofal, D. Spatial Analysis for the Social Sciences; Cambridge University Press: New York City, NY, USA, 2015. [Google Scholar]
- Banerjee, S.; Wall, M.M.; Carlin, B.P. Frailty Modeling for Spatially Correlated Survival Data, with Application to Infant Mortality in Minnesota. Biostatistics
**2003**, 4, 123–142. [Google Scholar] [CrossRef] [PubMed] - Taylor, B.; Rowlingson, B. spatsurv: An R package for Bayesian inference with spatial survival models. J. Stat. Softw.
**2017**, 77, 1–32. [Google Scholar] [CrossRef] - Muff, S.; Signer, J.; Fieberg, J. Accounting for individual-specific variation in habitat-selection studies: Efficient estimation of mixed-effects models using Bayesian or frequentist computation. J. Anim. Ecol.
**2020**, 89, 80–92. [Google Scholar] [CrossRef] [PubMed] - Svolik, M. Authoritarian Reversals and Democratic Consolidation. Am. Political Sci. Rev.
**2008**, 102, 153–168. [Google Scholar] [CrossRef] - Walter, B.F. Why bad governance leads to repeat civil war. J. Confl. Resolut.
**2015**, 59, 1242–1272. [Google Scholar] [CrossRef] - Maller, R.A.; Zhou, X. Survival Analysis with Long-Term Survivors; Wiley: New York, NY, USA, 1996; Volume 525. [Google Scholar]
- Yin, G.; Ibrahim, J.G. Cure rate models: A unified approach. Can. J. Stat.
**2005**, 33, 559–570. [Google Scholar] [CrossRef] - Beger, A.; Hill, D.W.; Metternich, N.W.; Minhas, S.; Ward, M.D. Splitting it up: The spduration split-population duration regression package for time-varying covariates. R J.
**2017**, 9, 474–486. [Google Scholar] [CrossRef] - Lu, W. Efficient estimation for an accelerated failure time model with a cure fraction. Stat. Sin.
**2010**, 20, 661. [Google Scholar] [PubMed] - Peng, Y.; Taylor, J. Cure models. In Handbook of Survival Analysis; Klein, J.P., Van Houwelingen, H.C., Ibrahim, J.G., Scheike, T.H., Eds.; Chapman and Hall: Boca Raton, FL, USA, 2014; pp. 113–134. [Google Scholar]
- Ibrahim, J.G.; Chen, M.H.; Sinha, D. Criterion-based methods for Bayesian model assessment. Stat. Sin.
**2001**, 11, 419–443. [Google Scholar] - Bernardinelli, L.; Montomoli, C. Empirical Bayes versus fully Bayesian analysis of geographical variation in disease risk. Stat. Med.
**1992**, 11, 983–1007. [Google Scholar] [CrossRef] [PubMed] - Besag, J.; York, J.; Mollié, A. Bayesian Image Restoration, with Two Applications in Spatial Statistics. Ann. Inst. Stat. Math.
**1991**, 43, 1–20. [Google Scholar] [CrossRef] - Thomas, A.; Best, N.; Lunn, D.; Arnold, R.; Spiegelhalter, D. GeoBUGS User Manual, Version 1.2. Available online: http:www.mrc-bsu.cam.ac.uk/bugs/ (accessed on 15 January 2021).
- Geweke, J. Evaluating the Accuracy of Sampling-based Approaches to the Calculation of Posterior Moments. In Bayesian Statistics; Bernardo, J., Berger, J., Dawid, A., Smith, A., Eds.; Clarendon Press: Oxford, UK, 1992. [Google Scholar]
- Heidelberger, P.; Welch, P.D. Simulation run length control in the presence of an initial transient. Oper. Res.
**1983**, 3, 1109–1144. [Google Scholar] [CrossRef] - Carlin, B.; Louis, T. Bayes and Empirical Bayes Methods for Data Analysis; Chapman and Hall/CRC: New York, NY, USA, 2000. [Google Scholar]
- Neal, R.M. Slice sampling. Ann. Stat.
**2003**, 31, 705–767. [Google Scholar] [CrossRef] - Gasiorowski, M.J.; Power, M. The Structural Determinants of Democratic Consolidation. Comp. Political Stud.
**1998**, 31, 740–771. [Google Scholar] [CrossRef] - Gassebner, M.; Lamla, M.J.; Vreeland, J.R. Extreme Bounds of Democracy. J. Confl. Resolut.
**2013**, 57, 171–197. [Google Scholar] [CrossRef] - Cheibub, J.A. Presidentialism, Parliamentarism, and Democracy; University Press: Cambridge, NY, USA, 2007. [Google Scholar]
- Przeworski, A.; Alvarez, M.; Cheibub, J.A.; Limongi, F. Democracy and Development: Political Institutions and Economic Performance, 1950–1999; Cambridge University: New York City, NY, 2000. [Google Scholar]
- Treisman, D. Is Democracy in Danger? A Quick Look at the Data. Prepared for the Conference on “Democratic Backsliding and Electoral Authoritarianism. Available online: https://www.danieltreisman.org/s/draft-june-7.pdf (accessed on 8 November 2018).
- Boix, C.; Stokes, S.C. Endogenous democratization. World Politics
**2003**, 55, 517–549. [Google Scholar] [CrossRef] - Przeworski, A.; Limongi, F. Modernization: Theories and facts. World Politics
**1997**, 49, 155–183. [Google Scholar] [CrossRef] - Gleditsch, K.S.; Ward, M.D. Diffusion and the International Context of Democratization. Int. Organ.
**2006**, 60, 911–933. [Google Scholar] [CrossRef] - Kopstein, J.S.; Reilly, D.A. Geographic Diffusion and the Transformation of the Postcommunist World. World Politics
**2000**, 53, 1–37. [Google Scholar] [CrossRef] - Brinks, D.; Coppedge, M. Diffusion is no illusion: Neighbor emulation in the third wave of democracy. Comp. Political Stud.
**2006**, 39, 463–489. [Google Scholar] [CrossRef] - Cliff, A.D.; Ord, J.K. Spatial Processes: Models & Applications; Pion: London, UK, 1981. [Google Scholar]
- Moran, P.A. Notes on continuous stochastic phenomena. Biometrika
**1950**, 37, 17–23. [Google Scholar] [CrossRef] [PubMed] - Toft, M.D. Ending civil wars: A case for rebel victory? Int. Secur.
**2010**, 34, 7–36. [Google Scholar] [CrossRef] - Licklider, R. The consequences of negotiated settlements in civil wars, 1945–1993. Am. Political Sci. Rev.
**1995**, 89, 681–690. [Google Scholar] [CrossRef] - Fortna, V.P. Does peacekeeping keep peace? International intervention and the duration of peace after civil war. Int. Stud. Q.
**2004**, 48, 269–292. [Google Scholar] [CrossRef] - Walter, B.F. Does conflict beget conflict? Explaining recurring civil war. J. Peace Res.
**2004**, 41, 371–388. [Google Scholar] [CrossRef] - Gates, S.; Graham, B.A.; Lupu, Y.; Strand, H.; Strøm, K.W. Power sharing, protection, and peace. J. Politics
**2016**, 78, 512–526. [Google Scholar] [CrossRef] - Braithwaite, A. Resisting infection: How state capacity conditions conflict contagion. J. Peace Res.
**2010**, 47, 311–319. [Google Scholar] [CrossRef] - Buhaug, H.; Gleditsch, K.S. Contagion or confusion? Why conflicts cluster in space. Int. Stud. Q.
**2008**, 52, 215–233. [Google Scholar] [CrossRef] - Gates, S.; Hegre, H.; Jones, M.P.; Strand, H. Institutional inconsistency and political instability: Polity duration, 1800–2000. Am. J. Political Sci.
**2006**, 50, 893–908. [Google Scholar] [CrossRef] - Elbadawi, E.; Sambanis, N. Why are there so many civil wars in Africa? Understanding and preventing violent conflict. J. Afr. Econ.
**2000**, 9, 244–269. [Google Scholar] [CrossRef] - Doyle, M.W.; Sambanis, N. International peacebuilding: A theoretical and quantitative analysis. Am. Political Sci. Rev.
**2000**, 94, 779–801. [Google Scholar] [CrossRef] - Murdoch, J.C.; Sandler, T. Civil wars and economic growth: Spatial dispersion. Am. J. Political Sci.
**2004**, 48, 138–151. [Google Scholar] [CrossRef] - Collier, P.; Elliot, V.; Hegre, H.; Hoeffler, A.; Reynal-Querol, M.; Sambanis, N. Breaking the Conflict Trap: Civil War and Development Policy; The World Bank: Washington, DC, USA, 2003. [Google Scholar]
- Walter, B.F. Designing transitions from civil war: Demobilization, democratization, and commitments to peace. Int. Secur.
**1999**, 24, 127–155. [Google Scholar] [CrossRef] - Hays, J.C.; Schilling, E.U.; Boehmke, F.J. Accounting for right censoring in interdependent duration analysis. Political Anal.
**2015**, 23, 400–414. [Google Scholar] [CrossRef] - Beck, N.; Gleditsch, K.S.; Beardsley, K. Space is more than geography: Using spatial econometrics in the study of political economy. Int. Stud. Q.
**2006**, 50, 27–44. [Google Scholar] [CrossRef] - Franzese, R.J.; Hays, J.C. Spatial Econometric Models of Cross-Sectional Interdependence in Political Science Panel and Time-Series-Cross-Section Data. Political Anal.
**2007**, 15, 140–164. [Google Scholar] [CrossRef] - Wagner, H. Bayesian estimation and stochastic model specification search for dynamic survival models. Stat. Comput.
**2011**, 21, 231–246. [Google Scholar] [CrossRef] - Sparapani, R.A.; Logan, B.R.; McCulloch, R.E.; Laud, P.W. Nonparametric survival analysis using Bayesian additive regression trees (BART). Stat. Med.
**2016**, 35, 2741–2753. [Google Scholar] [CrossRef] - Rizki, S.W.; Mara, M.N.; Sulistianingsih, E. Survival bayesian estimation of exponential-gamma under LINEX loss function. Proc. J. Phys. Conf. Ser. IOP
**2017**, 855, 012036. [Google Scholar] [CrossRef] - Mahdizadeh, M.; Zamanzade, E. Goodness-of-fit testing for the Cauchy distribution with application to financial modeling. J. King Saud-Univ.-Sci.
**2019**, 31, 1167–1174. [Google Scholar] [CrossRef]

**Figure 1.**MC Experiment 1 $\beta $,$\gamma $ densities for SP Weibull, NS Frailty Weibull, and Spatial SP Weibull models for: (

**a**) $\widehat{{\beta}_{0}},N=100$, (

**b**) $\widehat{{\beta}_{0}},N=400$, (

**c**) $\widehat{{\beta}_{0}},N=1000$, (

**d**) $\widehat{{\beta}_{0}},N=1500$, (

**e**) $\widehat{{\beta}_{0}},N=2000$, (

**f**) $\widehat{{\beta}_{1}},N=100$, (

**g**) $\widehat{{\beta}_{1}},N=400$, (

**h**) $\widehat{{\beta}_{1}},N=1000$, (

**i**) $\widehat{{\beta}_{1}},N=1500$, (

**j**) $\widehat{{\beta}_{1}},N=2000$, (

**k**) $\widehat{{\gamma}_{0}},N=100$, (

**l**) $\widehat{{\gamma}_{0}},N=400$, (

**m**) $\widehat{{\gamma}_{0}},N=1000$, (

**n**) $\widehat{{\gamma}_{0}},N=1500$, (

**o**) $\widehat{{\gamma}_{0}},N=2000$, (

**p**) $\widehat{{\gamma}_{1}},N=100$, (

**q**) $\widehat{{\gamma}_{1}}$, $N=400$, (

**r**) $\widehat{{\gamma}_{1}},N=1000$, (

**s**) $\widehat{{\gamma}_{1}},N=1500$, (

**t**) $\widehat{{\gamma}_{1}},N=2000$, (

**u**) $\widehat{{\gamma}_{2}},N=100$, (

**v**) $\widehat{{\gamma}_{2}},N=400$, (

**w**) $\widehat{{\gamma}_{2}},N=1000$, (

**x**) $\widehat{{\gamma}_{2}},N=1500$, (

**y**) $\widehat{{\gamma}_{2}},N=2000$.

**Figure 2.**MC Experiment 2 mean RMSE comparison between SP Weibull, NS Frailty SP Weibull, and spatial SP Weibull models for (

**a**) $\widehat{\beta}$ coefficients and (

**b**) $\widehat{\gamma}$ coefficients with spatial dependence changing from 30% to 80% of the data.

**Figure 3.**MC Experiment 3 mean RMSE comparison between SP Weibull, NS Frailty SP Weibull, and spatial SP Weibull models for (

**a**) $\widehat{\beta}$ coefficients and (

**b**) $\widehat{\gamma}$ coefficients with the immune fraction changing from 25% to 60% of the data.

**Figure 4.**Democratic survival application spatial frailty maps: (

**a**) depicts the posterior mean estimates of

**V**(split−stage spatial frailties) and (

**b**) depicts the posterior mean estimates of

**W**(survival−stage spatial frailties).

**Figure 5.**Democratic consolidation stage ($\widehat{\gamma}$) coefficient results from SP Weibull, NS Frailty SP Weibull, and spatial SP Weibull models for the following covariates: (

**a**) GDP/cap, (

**b**) GDP growth, (

**c**) military government, (

**d**) monarchy, (

**e**) civilian government, (

**f**) parliamentary government, and (

**g**) presidential government.

**Figure 6.**Democratic survival stage ($\widehat{\beta}$) coefficient results from SP Weibull, NS Frailty SP Weibull, and Spatial SP Weibull models for the following covariates: (

**a**) GDP/cap, (

**b**) GDP growth, (

**c**) military government, (

**d**) monarchy, (

**e**) civilian government, (

**f**) parliamentary government, and (

**g**) presidential government.

**Figure 7.**Post−war peace duration application spatial frailty maps: (

**a**) depicts the posterior mean estimates of

**V**(split−stage spatial frailties) and (

**b**) depicts the posterior mean estimates of

**W**(survival−stage spatial frailties).

**Figure 8.**Peace consolidation ($\widehat{\gamma}$) coefficient results from SP Weibull and Spatial SP Weibull models for the following covariates: (

**a**) press freedom, (

**b**) victory, (

**c**) mountains, and (

**d**) GDP/cap.

**Figure 9.**Peace survival stage ($\widehat{\beta}$) coefficient results from SP Weibull and Spatial SP Weibull models for the following covariates: (

**a**) press freedom, (

**b**) GDP/cap, (

**c**) peace agreement, (

**d**) intensity, (

**e**) ethnic factionalization, (

**f**) UN peacekeeping, (

**g**) territory, (

**h**) non-contiguous, and (

**i**) mountains.

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |

© 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Joo, M.M.; Bolte, B.; Huynh, N.; Mukherjee, B.
Bayesian Spatial Split-Population Survival Model with Applications to Democratic Regime Failure and Civil War Recurrence. *Mathematics* **2023**, *11*, 1886.
https://doi.org/10.3390/math11081886

**AMA Style**

Joo MM, Bolte B, Huynh N, Mukherjee B.
Bayesian Spatial Split-Population Survival Model with Applications to Democratic Regime Failure and Civil War Recurrence. *Mathematics*. 2023; 11(8):1886.
https://doi.org/10.3390/math11081886

**Chicago/Turabian Style**

Joo, Minnie M., Brandon Bolte, Nguyen Huynh, and Bumba Mukherjee.
2023. "Bayesian Spatial Split-Population Survival Model with Applications to Democratic Regime Failure and Civil War Recurrence" *Mathematics* 11, no. 8: 1886.
https://doi.org/10.3390/math11081886