Exogenous variation

More money to fund institutions two things that

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: hese things correlated does not necessarily mean the arrow runs this way) Some third factor is causing both prosperity and institutional quality In fact, development may cause institutions to improve; poverty may cause bad institutions. More money to fund institutions Two things that matter for property rights: Enforcement Information about property rights In richer societies, they may be better able to choose those institutions that maximize social surplus (enforcement, monitoring capacity) In poorer societies, institutions may be inefficient Or: development and institutions may be caused by the same things: Geography Historical conditions So insitutions may ACTUALLY BE ENDOGENOUS How to control for endogeneity? The best way to deal with endogeneity concerns is through instrumental variables techniques Find a genuinely exogenous variable that is strongly CORRELATED with the potentially endogenous regressor but does not affect the outcome, EXCEPT through this channel Z ! X ! Y only in this way. Y=alpha+beta(x)+epsilon Alpha is y- int; beta is slope, and epsilon is error (linear graph) A primer on IV regression: Step 1: X = a0 + a1Z1 + a2Z2 + akZ...
View Full Document

{[ snackBarMessage ]}

Ask a homework question - tutors are online