Consumption (a)symmetric response to predictable income: a new test for myopia and liquidity constraints

This paper examines whether consumption responds symmetrically or asymmetrically to predictable income. To accomplish such task, we propose an adjustment in Shea [1995a] testing equation that make straightforward to test whether consumption is more sensible to predictable income increases than decreases. Furthermore, our approach allows us to employ usual instrumental variable estimators and econometric tools developed to deal with the weak instruments problem. Our new approach yields the following results: i) there is overwhelming evidence that instruments are weak; ii) the point estimates for negative income growth are higher than those for positive income growth; iii) hypothesis testings indicate the same findings; iv) confidence sets robust to weak instruments show support for previous point estimates and hypothesis tests results. Therefore, instead of finding evidence for myopia or liquidity constraints, the findings support the “perverse asymmetry” hypothesis raised by Shea [1995a].