Statistical Science

Instrumental Variable Estimation with a Stochastic Monotonicity Assumption

Dylan S. Small, Zhiqiang Tan, Roland R. Ramsahai, Scott A. Lorch, and M. Alan Brookhart

Full-text: Access denied (no subscription detected)

We're sorry, but we are unable to provide you with the full text of this article because we are not able to identify you as a subscriber. If you have a personal subscription to this journal, then please login. If you are already logged in, then you may need to update your profile to register your subscription. Read more about accessing full-text


The instrumental variables (IV) method provides a way to estimate the causal effect of a treatment when there are unmeasured confounding variables. The method requires a valid IV, a variable that is independent of the unmeasured confounding variables and is associated with the treatment but which has no effect on the outcome beyond its effect on the treatment. An additional assumption often made is deterministic monotonicity, which says that for each subject, the level of the treatment that a subject would take is a monotonic increasing function of the level of the IV. However, deterministic monotonicity is sometimes not realistic. We introduce a stochastic monotonicity assumption, a relaxation that only requires a monotonic increasing relationship to hold across subjects between the IV and the treatments conditionally on a set of (possibly unmeasured) covariates. We show that under stochastic monotonicity, the IV method identifies a weighted average of treatment effects with greater weight on subgroups of subjects on whom the IV has a stronger effect. We provide bounds on the global average treatment effect under stochastic monotonicity and a sensitivity analysis for violations of stochastic monotonicity. We apply the methods to a study of the effect of premature babies being delivered in a high technology neonatal intensive care unit (NICU) vs. a low technology unit.

Article information

Statist. Sci. Volume 32, Number 4 (2017), 561-579.

First available in Project Euclid: 28 November 2017

Permanent link to this document

Digital Object Identifier

Causal inference observational study instrumental variable two stage least squares


Small, Dylan S.; Tan, Zhiqiang; Ramsahai, Roland R.; Lorch, Scott A.; Brookhart, M. Alan. Instrumental Variable Estimation with a Stochastic Monotonicity Assumption. Statist. Sci. 32 (2017), no. 4, 561--579. doi:10.1214/17-STS623.

Export citation


  • Abadie, A. (2003). Semiparametric instrumental variable estimation of treatment response models.J. Econometrics113231–263.
  • Angrist, J. D., Imbens, G. W. and Rubin, D. B. (1996). Identification of causal effects using instrumental variables.J. Amer. Statist. Assoc.91444–455.
  • Angrist, J. D. and Pischke, J.-S. (2009).Mostly Harmless Econometrics:An Empiricist’s Companion. Princeton Univ. Press, Princeton, NJ.
  • Bagnoli, M. and Bergstrom, T. (2005). Log-concave probability and its applications.Econom. Theory26445–469.
  • Baiocchi, M., Small, D. S., Lorch, S. and Rosenbaum, P. R. (2010). Building a stronger instrument in an observational study of perinatal care for premature infants.J. Amer. Statist. Assoc.1051285–1296.
  • Balke, A. and Pearl, J. (1997). Bounds on treatment effects for studies with imperfect compliance.J. Amer. Statist. Assoc.921171–1176.
  • Boef, A. G. C., Gussekloo, J., Dekkers, O. M., Frey, P., Kearney, P. M., Kerse, N., Mallen, C. D., McCarthy, V. J. C., Mooijaart, S. P., Muth, C., Rodondi, N., Rosemann, T., Russell, A., Schers, H., Virgini, V., de Waal, M. W. M., Warner, A., le Cessie, S. and den Elzen, W. P. J. (2016). Physician’s prescribing preference as an instrumental variable: Exploring assumptions using survey data.Epidemiology27276–283.
  • Brookhart, M. A. and Schneeweiss, S. (2007). Preference-based instrumental variable methods for the estimation of treatment effects: Assessing validity and interpreting results.Int. J. Biostat.3Art. 14, 25.
  • Brookhart, M. A., Wang, P., Solomon, D. H. and Schneeweiss, S. (2006). Evaluating short-term drug effects using a physician-specific prescribing preference as an instrumental variable.Epidemiology17268–275.
  • Cavolina, M. J. F., Kelly, M. A. T., Stone, J. A. J. and Davis, R. G. M. (2000).Growing up Catholic:The Millennium Edition:An Infinitely Funny Guide for the Faithful,the Fallen and Everyone in-Between. Image.
  • Chalak, K. (2017). Instrumental variables methods with heterogeneity and mismeasured instruments.Econometric Theory3369–104.
  • Cheng, J. and Small, D. S. (2006). Bounds on causal effects in three-arm trials with non-compliance.J. R. Stat. Soc. Ser. B Stat. Methodol.68815–836.
  • Cox, D. R. (1958).Planning of Experiments. Wiley, New York.
  • de Chaisemartin, C. (2017). Tolerating defiance: Local average treatment effects without monotonicity.Quant. Econ.8367–396.
  • Deaton, A. (2010). Instruments, randomization and learning about development.J. Econ. Lit.48424–455.
  • Dijkstra, E. W. (1959). A note on two problems in connexion with graphs.Numer. Math.1269–271.
  • DiNardo, J. and Lee, D. S. (2011). Program evaluation and research designs. InHandbook of Labor Economics,Vol.4 463–536. Elsevier, Amsterdam.
  • Finkelstein, A., Taubman, S., Wright, B., Bernstein, M., Gruber, J., Newhouse, J. P., Allen, H., Baicker, K. and Group Oregon Health Study (2012). The Oregon health insurance experiment: Evidence from the first year.Q. J. Econ.1271057–1106.
  • Guo, Z., Cheng, J., Lorch, S. A. and Small, D. S. (2014). Using an instrumental variable to test for unmeasured confounding.Stat. Med.333528–3546.
  • Hernán, M. A. and Robins, J. M. (2006). Instruments for causal inference: An epidemiologist’s dream?Epidemiology17360–372.
  • Horowitz, J. L. and Manski, C. F. (2000). Nonparametric analysis of randomized experiments with missing covariate and outcome data.J. Amer. Statist. Assoc.9577–88.
  • Huber, M. and Mellace, G. (2012). Relaxing monotonicity in the identification of local average treatment effects. Working paper.
  • Imbens, G. W. (2014). Instrumental variables: An econometrician’s perspective.Statist. Sci.29323–358.
  • Imbens, G. W. and Angrist, J. D. (1994). Identification and estimation of local average treatment effects.Econometrica61467–476.
  • Klein, T. J. (2010). Heterogeneous treatment effects: Instrumental variables without monotonicity?J. Econometrics15599–116.
  • Korn, E. L. and Baumrind, S. (1998). Clinician preferences and the estimation of causal treatment differences.Statist. Sci.13209–235.
  • Lehmann, E. L. (1966). Some concepts of dependence.Ann. Math. Statist.371137–1153.
  • Lorch, S. A., Baiocchi, M., Ahlberg, C. E. and Small, D. S. (2012a). The differential impact of delivery hospital on the outcomes of premature infants.Pediatrics130270–278.
  • Lorch, S. A., Kroelinger, C. D., Ahlberg, C. and Barfield, W. D. (2012b). Factors that mediate racial/ethnic disparities in US fetal death rates.Am. J. Publ. Health1021902–1910.
  • Manski, C. F. (1990). Non-parametric bounds on treatment effects.Am. Econ. Rev.80351–374.
  • McClellan, M., McNeil, B. J. and Newhouse, J. P. (1994). Does more intensive treatment of acute myocardial infarction in the elderly reduce mortality? Analysis using instrumental variables.J. Am. Med. Dir. Assoc.272859–866.
  • Neal, D. (1997). The effects of Catholic secondary schooling on educational achievement.J. Labor Econ.1498–123.
  • Ogburn, E. L., Rotnitzky, A. and Robins, J. M. (2015). Doubly robust estimation of the local average treatment effect curve.J. R. Stat. Soc. Ser. B. Stat. Methodol.77373–396.
  • Pearl, J. (2011). Principal stratification—a goal or a tool?Int. J. Biostat.7Art. 20, 15.
  • Phibbs, C. S., Mark, D. H., Luft, H. S., Peltzman-Rennie, D. J., Garnick, D. W., Lichtenberg, E. and McPhee, S. J. (1993). Choice of hospital for delivery: A comparison of high-risk and low-risk women.Health Serv. Res.28201.
  • Ramsahai, R. R. (2012). Causal bounds and observable constraints for non-deterministic models.J. Mach. Learn. Res.13829–848.
  • Robins, J. M. (1989). The analysis of randomized and nonrandomized AIDS treatment trials using a new approach to causal inference in longitudinal studies. InHealth Service Research Methodology:A Focus on AIDS(L. Sechrest, H. Freeman and A. Mulley, eds.) 113–159.
  • Robins, J. M. (1994). Correcting for non-compliance in randomized trials using structural nested mean models.Comm. Statist. Theory Methods232379–2412.
  • Rosenbaum, P. R. (2010).Design of Observational Studies. Springer, New York.
  • Rubin, D. B. (1986). Statistics and causal inference: Comment: Which ifs have causal answers.J. Amer. Statist. Assoc.81961–962.
  • Small, D. S., Tan, Z., Ramsahai, R. R., Lorch, S. A. and Brookhart, M. A. (2017). Supplement to “Instrumental variable estimation with a stochastic monotonicity assumption.”DOI:10.1214/17-STS623SUPP.
  • Stock, J. H. (2001). Instrumental variables in economics and statistics. InInternational Encyclopedia of the Social & Behavioral Sciences(N. J. Smelser and P. B. Baltes, eds.) 7577–7582. Elsevier, Amsterdam.
  • Swanson, S. A. and Hernán, M. A. (2014). Think globally, act globally: An epidemiologist’s perspective on instrumental variable estimation [discussion of MR3264545].Statist. Sci.29371–374.
  • Swanson, S. A., Miller, M., Robins, J. M. and Hernán, M. A. (2015). Definition and evaluation of the monotonicity condition for preference-based instruments.Epidemiology26414–420.
  • Tan, Z. (2006). Regression and weighting methods for causal inference using instrumental variables.J. Amer. Statist. Assoc.1011607–1618.
  • Tan, Z. (2010). Marginal and nested structural models using instrumental variables.J. Amer. Statist. Assoc.105157–169.
  • VanderWeele, T. J. and Shpitser, I. (2013). On the definition of a confounder.Ann. Statist.41196–220.
  • Yang, F., Lorch, S. A. and Small, D. S. (2014). Estimation of causal effects using instrumental variables with nonignorable missing covariates: Application to effect of type of delivery NICU on premature infants.Ann. Appl. Stat.848–73.

Supplemental materials

  • Supplement to “Instrumental Variable Estimation with a Stochastic Monotonicity Assumption”. Section A of the supplementary materials presents analogues of Propositions 1–4 when conditioning on observed covariates. Section B presents proofs for the propositions given in Section A of the supplementary materials. Section C presents proofs for the results in Section 2 of the main text. Section D shows that deterministic compliance class framework identification results are a special case of stochastic compliance class framework results. Section E discusses bounds on the global average treatment effect for a binary outcome under stochastic monotonicity. Section F gives an example in which bounds under stochastic monotonicity are tighter than bounds without stochastic monotonicity. Section G extends results in the main text to the setting of a non-binary instrumental variable. Section H discusses identification results when the measured IV is an intensity preserving proxy for a continuous causal IV satisfying deterministic monotonicity. Sections I and J give additional examples besides those given in the main text in which deterministic monotonicity is likely violated but stochastic monotonicity is plausible.