separate questions on life satisfaction and happiness.5 (Since the WVS does not contain
questions on job satisfaction, our paper does not deal with Blanchflower and Oswald’s
findings regarding this issue.)
The main disadvantage of working with the WVS data is that the data are
comprised of two separate survey waves. Of the 78 countries, 12 (including Australia)
were surveyed in 1995-97, while a further 66 were surveyed in 1999-2000. Where
countries were surveyed in both waves, only the later survey is included in the integrated
data file. We checked other surveys, and confirmed that Australian levels of happiness
and life satisfaction have been stable over the past decade.6 Comparing Australia in 1995-
97 with other countries in 1999-2000 is therefore unlikely to bias our results.
For consistency with the previous section, we continue to compare these
outcomes with the Human Development Index as measured in 2002, although given the
persistence in the HDI (and wellbeing indicators), this does not much affect our findings.
Following Blanchflower and Oswald, we simply use each country’s mean response to the
wellbeing questions