High positive correlation between input and efficiency severely biases performance.
The article discusses how endogeneity can affect the accuracy of efficiency estimates in economic models. By using data envelopment analysis (DEA), the researchers found that high positive endogeneity levels can significantly bias DEA performance. This means that when one input is strongly correlated with the efficiency level, it can distort the results of DEA calculations. While DEA is resilient to negative endogeneity, positive endogeneity can greatly impact its accuracy.