Instrumental Variables Regressions with Honestly Uncertain Exclusion Restrictions

The validity of instrumental variables (IV) regression models depends crucially on fundamentally untestable exclusion restrictions. Typically exclusion restrictions are assumed to hold exactly in the relevant population, yet in many empirical appli...

Full description

Bibliographic Details
Main Author:	Kraay, Aart
Format:	Policy Research Working Paper
Language:	English
Published:	World Bank, Washington, DC 2012
Subjects:	ABSOLUTE VALUE BAYESIAN ANALYSIS BENCHMARK BILATERAL TRADE CAUSATION CONFIDENCE INTERVALS CONSUMERS CONTROL VARIABLES CORRELATIONS COUNTRY DUMMIES COVARIANCE DEGREES OF FREEDOM DEPENDENT VARIABLE DISTRIBUTIONAL ASSUMPTIONS DUMMY VARIABLE ECONOMETRIC THEORY ECONOMIC ACTIVITY ECONOMIC DEVELOPMENT ECONOMIC GROWTH ENDOGENOUS VARIABLE ENDOGENOUS VARIABLES EQUATIONS ERROR ERROR TERM ERROR TERMS EXCLUSION EXPERIMENTAL DATA FINANCIAL DEVELOPMENT FINANCIAL SERVICES FITTED VALUES GAMMA DISTRIBUTION GDP GDP PER CAPITA GROWTH RATE GROWTH RATES HUMAN CAPITAL INFLATION INSTRUMENTAL VARIABLE INSTRUMENTAL VARIABLES INSTRUMENTAL VARIABLES ESTIMATOR INSTRUMENTAL VARIABLES REGRESSION INTERNATIONAL BANK LIKELIHOOD FUNCTION LINEAR FUNCTION LINEAR REGRESSION LINEAR REGRESSION MODEL MACROECONOMICS MATRIX MORTALITY NORMAL DISTRIBUTION 0 HYPOTHESIS PER CAPITA INCOMES PRECISION PROBABILITY PROBABILITY DISTRIBUTION PROPERTY RIGHTS RANDOM VARIABLE RANDOM VARIABLES REGRESSION MODEL ROBUSTNESS CHECKS SAMPLE MEAN SAMPLE SIZE SENSITIVITY ANALYSIS SLOPE COEFFICIENT SLOPE COEFFICIENTS STANDARD DEVIATION STANDARD ERRORS STRUCTURAL PARAMETERS TEST STATISTICS TRADE EQUATION TRADE OPENNESS TRADE POLICY TRADE SHARE UNCERTAINTY UNKNOWN PARAMETER VALIDITY VARIABLE ESTIMATION VARIANCE-COVARIANCE MATRIX
Online Access:	http://documents.worldbank.org/curated/en/2008/05/9473668/instrumental-variables-regressions-honestly-uncertain-exclusion-restrictions http://hdl.handle.net/10986/6693

Description
Summary:	The validity of instrumental variables (IV) regression models depends crucially on fundamentally untestable exclusion restrictions. Typically exclusion restrictions are assumed to hold exactly in the relevant population, yet in many empirical applications there are reasonable prior grounds to doubt their literal truth. In this paper I show how to incorporate prior uncertainty about the validity of the exclusion restriction into linear IV models, and explore the consequences for inference. In particular I provide a mapping from prior uncertainty about the exclusion restriction into increased uncertainty about parameters of interest. Moderate prior uncertainty about exclusion restrictions can lead to a substantial loss of precision in estimates of structural parameters. This loss of precision is relatively more important in situations where IV estimates appear to be more precise, for example in larger samples or with stronger instruments. The author illustrates these points using several prominent recent empirical papers that use linear IV models.

Instrumental Variables Regressions with Honestly Uncertain Exclusion Restrictions

Similar Items