Alternative Specification Sample Clauses

Alternative Specification. A potential issue with the main specification could arise if some of those individuals who are classified treated as a result of having benefitted from spillover effects of the reform received their raise because they are more productive or even in better health. In an additional specification, I follow the approach used by Xxxxxxx and Xxxxxxxxx (2002) and construct an hourly wage measure for the pre-treatment period by using reported monthly income and hours worked per week. Again, the sample is restricted to low-wage workers: Individuals earning below the NMW in the year before the reform are considered as treated, whereas those earning above the NMW and below £6.00 form the control group.

Alternative Specification. Table 2.7 presents estimates obtained from an alternative DD specification, which uses a constructed measure of hourly wages in order to assign workers into treatment and control group. It is noticeable that the ordered logit results remain almost unchanged from the baseline estimates shown in Table 2.4. Receiving an increase in wages following the reform increases the likelihood of reporting excellent health by 3.88 percentage points. The estimate is statistically significant at the 1% level. The consistency in the finding across both specifications removes concerns regarding the fact that some workers who are classified as treated in the baseline model are receiving a raise due to higher productivity.