One potential problem is that consumption and SWB might be jointly determined. This would make consumption endogenous in equation (7), which generally leads to inconsistency. Put differently, it is possible that people with higher SWB behave differently than people with lower SWB. Guven (2012) find that ignoring simultaneity of happiness and consumption leads to downward bias. This implies that the causal effect is, if anything, stronger than the correlative coefficient suggests.23