Income Mobility of Owners of Small Businesses when Boundaries between Occupations are Vague

Regarding the relationship to other variables, we see that being male, having children, living
in an urban area, and length of education are positively correlated with upward mobility, whereas age,
being married and birth outside Western Europe, North-America and Oceania are not. The dummy
variables for field of education are defined in relation to the omitted reference category, an unspecified
broad field of education. Note also that the likelihood ratio test of a non-significant random effect is
clearly rejected; the random effects model is therefore preferable to alternatives without individual
effects, such as pooled probit. However, as noted earlier, the random effects estimation involves non-
trivial restrictions in the sample employed in the estimations, i.e. restricting to individuals represented
in more than one wave. Results from pooled probit estimations are shown in Table A1 in the
Appendix.

As we see, the parameter estimates for “male” are just below the parameter estimates for
small business ownership. Another way of setting parameter estimates for business ownership into
perspective is to compare them to a similar random effects estimation of downward mobility, i.e.,
using the same explanatory variables for transitions from quintiles 2-5 down to quintile 1. Such
estimation generates (positive) parameter values for business ownership of approximately the same
size as those featured in Table 10. The present analysis can therefore be said to confirm the uncertain
character of income from business ownership, leading to overrepresentation of owners of small
businesses at the low and high ends of the income distribution scale, see Table 2 and Table 3.

4.3.2 The relationship to staying rich

The next question concerns the extent to which involvement in business ownership increases the
probability of staying at the high end of the income distribution scale. Table 11 shows results of
random effects probit estimations based on conditional probabilities for staying in the upper quintile or
in the upper decile in year t, given that the persons were in the same position in year t-1. As we see,
the relationship is negative according to the quintile specification and positive only after restricting to
rich in terms as defined by the decile specification. However, again the parameter estimates are small.
The probability of staying rich is higher if the subject is male, married, has children, lives in an urban
setting, was born in Western Europe, North America or Oceania, and is well educated. Most
importantly, the wider definition of income strengthens the positive relationship for the decile
specification and reduces the negative effect according to the quintile alternative.

As measures of upward mobility were compared to downward mobility in the previous
section, it is illustrative to compare the estimates of Table 11 to similar estimates from “stay-poor”
estimations. Such comparisons reveal that business ownership is positively related to staying at the
low end of the income distribution, again confirming the symmetry of relationships to business
ownership at both tails of the income distribution.

25

<<   23   24   25   26   27   28   29   >>

More intriguing information