The zip file IIVcode.zip contains matlab code to consistently estimate and perform inference on bounds produced using imperfect instruments as described in Nevo and Rosen (2012): "Identification with Imperfect Instruments." The file "IIV_Bounds.m" contains the main file that can be used to produce bounds and confidence interval estimates. The confidence intervals provide asymptotically valid coverage for the population bounds using Bonferroni's inequality. Inequality selection may be conducted as described in Nevo and Rosen (2012), based on the working paper Chernozhukov, Lee, and Rosen (2009), or using the adaptive inequality selection procedure from Chernozhukov, Lee, and Rosen (2013). The file "IIV_cereal_code_example_usage.m" demonstrates the use of the IIV_Bounds function. Comments in the Matlab code provide further details.
This code is freely provided without warranty under the terms of the MIT License of the Open Source Initiative.