Simon Burgess (University of Bristol)
Ellen Greaves (University of Bristol)
Richard Murphy (University of Texas at Austin)
What happens when you deregulate the teacher labour market? Following a dramatic policy change in England, we get a chance to find out. In September 2013 the Coalition Government ended the use of ‘annual progression’ pay scales in all schools in England and required (not just ‘permitted’) all 20,000 plus state-funded schools to introduce their own Performance Related Pay (PRP) scheme. In doing so, the reform completely changed the basis for setting pay, affecting the whole labour market of close to half a million teachers in the public sector. Our discussion of the details of the reform and schools’ initial responses is here. And despite the scale of the changes to pay regulations, there were no changes to the way schools were financed, which allows us to isolate the impact of pay rigidities.
So, when schools can set pay as they want, within a budget constraint, what do they do? Specifically, we use this reform to address three key questions: first, how do schools change pay when given the freedom to do so? Second, how do these decisions affect the number of teachers employed? Finally, and possibly most importantly, does the decentralization of pay affect pupil performance? Details of our answers are in our just-out report, and we summarise them here.
First, it isn’t straightforward to answer these issues; we can’t just compare before and after because this was a time when many other potentially relevant things were happening, not least a slew of other education reforms, as well as standard year-to-year changes in practice and procedures. To be able to identify the effect we have to look for a factor that might provoke differential responses. We use the fact that the old teacher pay scales did not really take into account variation in local wages across labour markets (just variations in London). So we would expect schools situated in competitive high wage areas to react differently to the reform than schools in low wage areas. The sign and magnitude of these differences informs us about the extent to which the old pay scales took schools away from their desired levels of pay and teacher employment.
We have data on all the teachers in state schools in England, over all the years since 2010. Using that, we create for each individual teacher the counterfactual expected pay growth, estimating what they would have earned under the old scale point system. This is a neat way of taking into account the demographic composition of each school; for example, how many long-serving (and therefore expensive) teachers each school has. This allows us to consider deviations of actual pay from this expected pay growth as a result of the reform. We average these teacher level measures for each school, and this is our measure of the impact of the reform.
We are now in a position to answer our three questions.
- Post reform there is a general decrease in teacher pay growth coinciding with central budget cut-backs. But within that overall average decrease, there were systematic and substantial differences. Schools used their new flexibility to respond to local market conditions. Specifically, teachers’ salaries grew relatively faster in high wage labour markets. For example, primary schools at the 75th rather than the 25th percentile of local wage distribution increase teacher salaries by an additional 0.43 percent points (£120) annually. This is exactly as you would expect, and allowed school salaries in high-wage labour markets to start becoming more competitive.
- Freed from the constraints of the pay scales, some schools were able to offer relatively higher salaries and become more attractive as a place to work. Such schools were therefore able to expand employment relatively more than those in low-wage areas; in other words, the actions of these schools were successful in attracting and retaining more teachers. (Note that we keep saying “relatively more”, because that is what we are talking about here – the differential effect of the reform in high- and low-wage areas; the overall absolute changes in wages and employment over this period were also affected by other factors, notably the central schools budget). Interestingly, we find no (relative) increase in the number of newly hired teachers, meaning that the increase in employment occurred through a reduction in the outflow of teachers: schools used their funds to retain their existing teachers.
- Our third finding is that schools in high-wage labour markets achieve relatively higher growth in student test scores. The gains in student test scores are larger in schools with more disadvantaged student populations. Moreover, as the majority of the test score gains occurred during the first year of the reform before pay or the quantity of teachers employed were able to change, implying that much of these initial gains are due to response from the incumbent teachers. For each pound that local average hourly wages are higher, primary schools post-reform experience a 1.4 percent of a standard deviation increase in student test scores; secondary schools in high-wage areas see a larger increase of 3.3 percent of a standard deviation.
Of course, there are benefits to having a defined pay scales, such as greater clarity and certainty on pay for teachers, and a clear way of minimizing favouritism or discrimination in pay. (Indeed, post-reform we did find a small relative increase in male teachers’ pay that is hard to interpret). But our research shows that there are also significant and substantial benefits to removing central pay scales. National pay scales prevent local managers from allocating resources efficiently. Given autonomy, schools in high wage areas depart from the salaries determined by the national pay scales in order to increase pay and pay dispersion. This has allowed them to retain experienced staff. Perhaps most importantly, we find evidence that productivity, measured by relative growth in pupil test score gains, improves once schools have more autonomy on pay.