ISER Working Paper Series 2006-52
Paid holiday entitlements, weekly working hours and earnings in the UK
30 Oct 2006
Full-time workers in the UK enjoy similar holiday entitlements on average to those in other western European countries (about 4-5 weeks per year) but there are much larger differences across jobs in the UK. Just before the introduction in 1998 of the Working Time Regulations (WTR), which imposed a minimum entitlement (initially) of 3 weeks, 27% of full-time workers received more than 5 weeks' holiday, while 12% got 3 weeks or less. This paper examines what can account for the dispersion of holiday entitlements and looks at the relationship of entitlements to earnings. Are holiday entitlements used to supplement earnings in remuneration packages, or do jobs with longer holiday entitlements pay less than jobs with more generous entitlements? And how does the association of earnings and holiday entitlements compare to the relationship between earnings and the other dimension of annual working time, weekly working hours?
Job characteristics like industry, occupation, firm size and job tenure all help to explain holiday entitlements, as do personal characteristics like occupation and education. Long entitlements are associated with high qualifications, long job tenure, managerial and professional occupations, as well as public sector and unionised jobs. But these factors cannot account for most of the differences, which implies there is a wide range of remuneration policies across employers or that some workers manage to obtain higher levels of holiday entitlement than other comparable workers.
Holidays are strongly associated with pay and tend to reinforce differences in earnings rather than compensating for them. Workers with more than 5 weeks' entitlements in 1997 were generally in the top quarter of the full-time earnings distribution, while those with less than three weeks' entitlement were in the bottom third of the distribution. The positive earnings-holidays relationship - or equivalently the negative relationship between earnings and the number of annual weeks worked in the job - remains after controlling for differences in qualifications, occupation, industry etc. After taking account of these factors, a week's extra holiday entitlement - corresponding to 2% fewer weeks worked per year - is associated with approximately 4% higher earnings. This result seemingly contradicts simple economic theory, which predicts that a competitive market should reward shorter entitlements with higher earnings. On the other hand, consistent with theory, the paper finds a positive relationship between earnings and the number of hours worked per week: an increase of 1% in weekly hours is associated with 0.8% higher earnings on average.
The contrasting results for weekly working hours and annual working weeks can be explained if there are unmeasured factors like individual skills or differences in job quality which increase both earnings and holiday entitlements, whilst being unrelated to weekly work hours. A possible mechanism is that weekly work hours are relatively fixed by the requirements of the job, whereas holiday entitlements are more flexible. So if firm wishes to attract more, or better, workers by combining higher pay with reduced working time, it could be cheaper to increase holiday entitlements (and earnings) instead of reducing weekly hours. Supplementary analysis in the paper finds that the positive relationship between holidays and earnings cannot be explained by unmeasured factors (such as innate ability) that do not change over time. Instead, the process is more dynamic: as individuals' labour market positions develop - because they accumulate new skills or work their way into better jobs - holidays increase in tandem with earnings. Consistent with this interpretation, those workers whose low entitlements were raised by 'outside forces' (the WTR) do not seem to have experienced a change in earnings.