Estimating probabilities, marginal effects and treatment effects in recursive bivariate probit models
https://doi.org/10.55959/MSU0130-0105-6-58-4-2
Abstract
This article addresses the interpreting results of recursive system of binary equations estimation in the case when the system does not satisfy exclusion restriction conditions. It means that the equation defining the endogenous variable does not contain a unique covariate. This article extends the analysis of previous studies on the identifiability of parameters of recursive binary systems by analyzing the conditions for the identifiability of probabilities, marginal effects and treatment effects. We provide a reasonable consideration suggesting that even if parameters of the model are unidentifiable, it is still possible to estimate accurately the conditional probabilities and marginal effects, but not the treatment effects. The problem of identifiability discussed in the paper is also considered on real data. We estimate the probability of purchasing medicine depending on individuals’ characteristics and the fact of visiting a doctor. An important practical contribution of the work is the recommendation for researchers to interpret the result recursive binary system estimation via marginal effects in the case when it is not possible to include at least one unique variable in the equation for the binary covariate.
Keywords
recursive bivariate probit model,
binary endogenous covariate,
exclusion restrictions,
marginal effects,
treatment effects
Supporting agencies: This work is supported by the Russian Science Foundation, RSF 20-18-00427.
About the Authors
E. V. Kossova
HSE University
Russian Federation
Russian Federation
Moscow
I. S. Slabolitskiy
HSE University
Russian Federation
Russian Federation
Moscow
B. S. Potanin
HSE University
Russian Federation
Russian Federation
Moscow
References
1. Zasimova, L., & Kossova, E. (2016). Empirical analysis of out-of-pocket expenditures on medicine in Russia. Applied Econometrics, 42, 75–99.
2. Ayalew, M., & Xianzhi, Z. (2019). The eff ect of fi nancial constraints on innovation in developing countries: Evidence from 11 African countries. Asian Review of Accounting, 28 (3), 273–308. https://doi.org/10.1108/ara-02-2019-0036
3. Blasch, J., Filippini, M., & Kumar, N. (2019). Boundedly rational consumers, energy and investment literacy, and the display of information on household appliances. Recourse and Energy Economics, 59, 39–58. https://doi.org/10.2139/ssrn.2789350
4. Castello, J. (2012). Promoting employment of disabled women in Spain; Evaluating a policy. Labour Economics, 19 (1), 82–91. https://doi.org/10.1016/j.labeco.2011.08.003
5. Cohen, J., Azarova, V., Kollmann, A., & Reichl, J. (2019). Q-complementarity in household adoption of photovoltaics and electricity-intensive goods: The case of electric vehicles. Energy Economics, 83, 567–577. https://doi.org/10.1016/j.eneco.2019.08.004
6. Dogbe, W. (2020). Can poverty status explain obesity in developing countries? Evidence from Ghana. Agribusiness, 37 (2), 409–421. https://doi.org/10.1002/agr.21663
7. Filippini, M., Greene, W., Kumar, N., & Martinez-Cruz, A. (2018). A note on the diff erent interpretation of the correlation parameters in the bivariate and the recursive bivariate probit. Economicas Letters, 167, 104–107. https://doi.org/10.2139/ssrn.3038804
8. Freedman, D., & Sekhon, J. (2020). Endogeneity in probit response model. Cambridge University Press, 18 (2), 138–150. https://doi.org/10.2139/ssrn.1138489
9. Frondel, M., Horbach, J., & Rennings, K. (2008). What triggers environmental management and innovation? Empirical evidence for Germany. Ecological Economics, 66 (1), 153–160. https://doi.org/10.2139/ssrn.556945
10. Han, S., & Lee, S. (2019). Estimation in a generalization of bivariate probit models with dummy endogenous regressors. Journal of Applied Econometrics, 34 (6), 994–1015. https://doi.org/10.2139/ssrn.3233422
11. Han, S., & Vytlacil, E. (2017). Identifi cation in a generalization of bivariate probit models with dummy endogenous regressors. Journal of Econometrics, 199 (1), 63–73. https://doi.org/10.1016/j.jeconom.2017.04.001
12. Heckman, J. (1978). Dummy endogenous variables in a simultaneous equation system. Econometrica, 46 (4), 931–959. https://doi.org/10.3386/w0177
13. Ma, W., Abdulai, A., & Goetz, R. (2017). Agricultural cooperative and investment in organic soil amendments and chemical fertilizer in China. American Journal of Agricultural Economics, 100 (2), 502–520. https://doi.org/10.1093/ajae/aax079
14. Maddala, G. (1983). Limited-depended and qualitative variables in econometrics. Cambridge University Press. https://doi.org/10.1017/ccol052124143x.005
15. Martin, E., & Quintata, F. (2022). Consistency and identifi ability revisited. Brazilian Journal of Probability and Statistics, 16, 99–106.
16. Morris, S. (2007). The impact of obesity on employment. Labour Economics, 14 (3), 413–433. https://doi.org/10.1016/j.labeco.2006.02.008
17. Mourifi e, I., & Meango, R. (2014). A note on the identifi cation in two equations probit model with dummy endogenous regressor. Economics Letters, 125 (3), 360–363. https://doi.org/10.2139/ssrn.2340187
18. Shaikh, A., & Vytlacil, E. (2011). Partial identifi cation in triangular systems of equations with binary dependent variabls. Econometrica, 79 (3), 949–955. https://doi.org/10.3982/ecta9082
19. Sirven, N., & Debrand, T. (2012). Social capital and health of older Europeans: Casual pathways and health inequalities. Social Science & Medicine, 75 (7), 1288–1295. https://doi.org/10.1016/j.socscimed.2012.05.009
20. Wilde, J. (2000). Identifi cation of multiple equation probit models with endogenous dummy regressor. Economics Letters, 69 (3), 309–312. https://doi.org/10.1016/s0165-1765(00)00320-7
Supplementary files
Review
For citations:
Kossova E.V., Slabolitskiy I.S., Potanin B.S. Estimating probabilities, marginal effects and treatment effects in recursive bivariate probit models. Moscow University Economics Bulletin. 2023;(4):23-49. (In Russ.) https://doi.org/10.55959/MSU0130-0105-6-58-4-2
Views:
436
Email this article 