Instrument-free inference under confined regressor endogeneity; derivations and applications

Stellenbosch Working Paper Series No. WP09/2020
 
Publication date: June 2020
 
Author(s):
Jan F. Kiviet (University of Amsterdam and Stellenbosch University)
 
Abstract:

Instead of exploiting instruments and claiming these to be uncorrelated with the disturbances, in an instrument-free approach one may adopt flexible bounds on the correlation between the endogenous regressors and the disturbances. Such an alternative to Two-Stage Least-Squares (TSLS) inference is developed here for general linear models with endogenous possibly time-dependent regressors. Earlier results enabled this just for rather restrictive mesokurtic i.i.d. cross-section data. In three empirical replication studies their underlying exclusion restrictions are shown to be most doubtful. Next, incredible (weak-instrument robust) TSLS inference is replaced by more reliable remarkably narrow instrument-free asymptotically conservative confidence intervals.

 
JEL Classification:

C12, C13, C21, C22, C26

Keywords:

endogeneity robust least-squares inference, new exclusion restrictions test, replication studies, sensitivity analysis of two-stage least-squares

Download: PDF (739 KB)

BER Weekly

25 Jul 2025 Budget hurdle cleared, but US tariff implementation remains a risk
It was another big week on the local political front, but with some constructive momentum. On the trade front, ahead of next week’s 1 August deadline, Trump announced another “massive” trade deal with Japan. The upcoming week is busy, with a slew of global and domestic data releases and several monetary policy decisions, including the SARB....

Read the full issue