Equation (1) is estimated in Table 4‘s first columns. Results indicate that, after controlling
a large set of personal characteristics, immigrants are more exposed to risk than natives.
Economic sector dummy variables are added in order to estimate Equation (2). Not surprisingly, model‘s explanatory power increases significantly. Evidently, an important part of risk exposure can be explained by economic sector. However, immigrant dummy
variable is still positive and significant. This means that even within economic sector, immigrants are more exposed to hazards than natives.
Regarding human capital variables, in both equations education significantly reduces worker‘s expected risks exposure. Workers that have secondary school education are less exposed to risks than those who only have primary school education. Nevertheless, both
groups are more exposed to hazards than those who received higher education. Job experience, measured in months, also reduces worker‘s expected exposure to risks.
Opposite, language constrains increases worker‘s expected exposure to risks, but its effects
are not statistically significant. Among personal characteristics, as expected, women are less exposed to risks than men. Age variables‘ coefficients are negative, suggesting that younger workers are more exposed to risk than older ones. The relation between family
structure and risk exposure is not as expected. Contrary to the literature, in the first equation single and widowed/divorced workers are less exposed to risks than married ones. When controlling for economic sector, family structure coefficients are not significant. Finally, in line with the literature, workers with a temporal contract are more exposed to risks.