Marital Status is Misunderstood in Happiness Models
One of the most common findings in contemporary empirical social science is that being married is associated with higher measured levels of happiness, or life satisfaction. The result seems to be consistent across both countries and time, and is apparently robust to statistical method, including with respect to econometric specification and fixed effects modelling. Our contribution is to propose that quality of a marriage is likely to be a very important factor in our understanding of the role of marital status, and to conjecture that for some married people being in an alternative state would be conducive to a higher level of happiness. We test this simple idea with conventional OLS modelling using life satisfaction data from three countries, the US, the UK and Germany, and the findings are very clear. We find that the coefficient on the marriage dummy is significant and important with the usual modelling but once marriage quality is controlled for, the effects of being married are extremely different between those in good compared to those in poor marriages. In all three data sets people in self-assessed poor marriages are fairly miserable, and much less happy than unmarried people, and people in self-assessed good marriages are even more happy than the literature reports. We also find that the results differ importantly between women and men, with members of the former sex showing a greater range of responses to marriage quality than do men. A final set of results is that, when marriage quality is controlled for, the apparent marriage effects on other outcome variables, such as self reported health and trust, change significantly.
|Date of creation:||14 May 2010|
|Contact details of provider:|| Postal: 221 Burwood Highway, Burwood 3125|
Phone: 61 3 9244 3815
Fax: +61 3 5227 2655
Web page: http://www.deakin.edu.au/buslaw/aef/index.php
More information through EDIRC