“All we want is a good local school”

Rebecca Allen and Simon Burgess

Two articles in the Times Education Supplement (TES) last Friday nicely illustrate the debate on school choice and school competition.

The first reports results from the British Social Attitudes Survey (BSAS), citing research by Sonia Exley, at the LSE, showing that most respondents thought that school choice was not a priority.

A familiar refrain in the school choice debate is that “all we want is a good local school”. There should be little doubt that this is indeed what most parents do want. We have used data from the Millennium Cohort Study to estimate the relative weights that parents place on the characteristics of primary schools. Unsurprisingly, school academic quality is positively valued, and distance between home and school is highly negatively valued. This makes a lot of sense: many parents have to make this journey four times a day. So, yes, people, do want a good local school.

But where does this take us? It is often said to imply that school choice is a distraction, an irrelevance. There is a side issue of whether choice is a good thing per se, as opposed to being functionally good. This is the thrust of the point above, that choice itself was not a priority, though the study also reports that 68% agreed that parents should have a basic right to choose their child’s school. Choice per se may become valuable once contrasted with the alternative of no choice.

But the main issue should be whether using school choice is a better way to allocate children to schools than alternatives. One alternative is implicit in the statement – children should go to their local school. In fact this gets a lot of support in the survey: the TES reports that 85% of respondents in the BSAS believe parents “should send their children to their local school”.

This idea would work well if families were not permitted to move house after the school admissions rule was changed. It is surely obvious why. We know that parents care a great deal about the school their child goes to. If the school allocation rule was simply “you will attend your local school”, then parents who were able to would ensure that their local school was the one they wanted by moving house.

It is quite possible that this would in fact lead to no less social segregation in schools, and almost certainly greater social segregation by neighbourhoods. While we found the relationship between school quality and moving house to be weaker than many might expect, this would undoubtedly be stronger in a world where your residence determined your school. It also does not do away with the concern about having to actually exercise a choice – it simply transfers it to a choice of neighbourhood and school combined.

So neighbourhood-based schooling would be very unlikely to resolve the issues of social segregation and choice-angst associated with choice-based schooling. It would also hand each school a local monopoly and, in the case of poorer families at least, a captive audience with no escape.

This connects to the second TES article, a leader on school competition. As the article notes, “Few things exercise critics of education policy more than the spectre of increased competition in our school system.” The argument balances the “un-school”-like, unorganised, chaotic and generally messy nature of competition with the potential for this to improve outcomes for students.

In fact, there is some evidence on this trade-off and what the net result of competition is (the article is mostly about competition for 6^th form entrants and allegations of mis-information, but the available evidence is about compulsory schooling).  While the international evidence is mixed, the UK evidence suggests that there is at the very best a weak and small positive effect of competition on student outcomes; a review is here. The interesting question is why competition doesn’t appear to do much. The answer appears to lie in market failures in the schools market. If these could be addressed, it may be that a competitive threat might do more to raise standards in poorly performing schools.

Much of the furore about school ‘choice’ or ‘competition’ is misplaced. It is not choice between schools per se, relative to other allocation rules, that causes the perceived unfairness. The focus for objections should be the way that places in over-subscribed schools are allocated. The proximity criterion – who lives closest gets in – is operated in almost every non-selective school. This directly relates the chances of getting in to the most popular schools to family income, damaging social mobility in a very clear way. If some or all places at an over-subscribed school were filled by a random ballot, then school choice would seem a very different beast.

Finally, the competition article talks of ruined lives: “If no authority oversees admissions, plots likely pupil numbers or configures special needs support, the results won’t just be missed targets or dicey operating margins, but ruined, real pupil lives.” It is also true that that poor communities trapped with low-performing schools ruins lives, that unaccountable and coasting schools also ruin lives. The debate is about how best to avoid ruined lives, not whether or not they should be ruined.

The finances of low income households

Sarah Smith

Recent news has drawn attention to a likely increase in the use of payday loans – short-term loans at staggeringly high rates of interest – to cover temporary income shortfalls, giving an insight into how low income households manage their finances during the economic downturn.

Another possibility – discussed in a recent paper – is that low-income households might buy a National Lottery ticket. Rather than being irrational, as is usually thought, this might actually be a reasonable choice for people who are faced with “lumpy” spending needs – such as replacing a consumer durable or paying off debts – and who have no savings and no access to credit on reasonable terms.

The basic argument is that a Lottery ticket gives someone the opportunity to forego a small amount of nondurable spending for the chance of a sizeable lump sum that could make them a lot better off  – for example by allowing them to buy a replacement washing machine or television.

Is this what people actually do? Well, perhaps it is no coincidence that the National Lottery operator Camelot recently announced its “highest-ever interim” lottery sales.

Rather than looking at Lottery sales, however, the paper looks at the question in a different way and asks whether sales of durable goods are more responsive to Lottery wins than to similar-sized cash windfalls from other sources – which would be consistent with people buying Lottery tickets as a way of financing durable goods. To take care of general differences in the effect of Lottery winnings (such as feeling lucky), the research compares the difference in response across two types of people – those who should otherwise be able to draw on savings and/or other forms of credit and those who cannot.

The findings provide quite a bit of support for the argument that at least some people might use the Lottery as a way of financing lumpy spending. Focusing on windfalls of between £200 and £5,000, purchases of consumer durables (fridges, washing machines, televisions etc) are shown to be significantly more responsive more to a Lottery win than to other types of windfall – and only for those who would otherwise have few alternative options. This is precisely what you would expect to find if the Lottery was being used to finance such purchases (and it is hard to think of another explanation for this result).

So, while the current economic downturn may bring bad news for most, it may well continue to be a good time for Camelot.

A Report of Two Halves

Simon Burgess

We published some research last Friday showing that students perform less well in their crucial GCSE exams in years when there is a major international football tournament taking place at the same time. For example, the FIFA World Cup in the summer of 2010, or the UEFA European Championship next summer, both overlap in part with the GCSE exam timetable.

With the draw for the groups in the European Championship taking place earlier that day, much of the comment naturally and sensibly focussed on the specific issue of the impact of next year’s tournament on exam scores. This is important: we estimate that the concurrence of the exams and saturation media coverage of the football reduces exam scores on average by around 0.12 standard deviations of pupil performance and by a lot more for some groups who reduce their effort a lot. These groups tend to be from poorer areas and predominantly (but by no means exclusively) male students.  Since these groups are already lower-performing groups, this means that education gaps will widen. We think of this impact arising through a reduction in student effort, with that time being spent instead on watching the football tournament. The variation in impact arises because of differing tastes for football, arising in turn from cultural norms and idiosyncratic factors, and from the differential effectiveness of an hour of study on exam performance.

However, there is also a broader significance to the research: finding that effort matters matters.

Recent research by economists has broadened out from the previous focus on cognitive ability, and a great deal of work has investigated the role of non-cognitive factors in educational attainment. Non-cognitive factors can be identified with personality traits (see Heckman), and one of the ‘big 5’ personality traits is ‘conscientiousness’, with the related traits of self-control, accepting delayed gratification, and a strong work ethic. Conscientiousness has been shown to be an excellent predictor of educational attainment and course grades. These aspects of self control and ability to concentrate are clearly related to the broad notion of effort we are using here. Our results on the importance of effort strengthen this evidence by isolating the effect of decisions on effort and time allocation in addition to the general ability to concentrate and exert self-control.

There is a great deal of policy interest in England arising from recent studies of US Charter schools with what is called a “No Excuses” ethos. This includes the KIPP (Knowledge Is Power Program) network of schools and schools in the Harlem Children’s Zone. These schools all feature a long school day, a longer school year, very selective teacher recruitment, strong norms of behaviour, as well as other characteristics. Some of the profession’s very top researchers have produced evidence showing that such schools produce very powerful positive effects on student achievement. While this overall effect could be due to different aspects of the KIPP/HCZ ethos, part of it is very likely to be increased effort from the students. Our results complement this by showing the impact of just a change in effort, and that that can have very substantial effects.

This matters for a number of reasons. First, unlike genetic characteristics, cognitive ability or non-cognitive traits, effort is almost immediately changeable. Our results suggest that this could have a big effect. The fact that we find changes in student effort to be very potent in affecting test scores suggests that policy levers to raise effort through incentives or changing school ethos are worth considering seriously. Such interventions would be justified if the low effort resulted from market failures due to lack of information on the returns to schooling, or time-inconsistent discounting.  Second, the importance of a manipulable factor such as effort for adolescents’ educational performance provides evidence of potentially high value policy interventions much later than “early years” policies. This is encouraging, offering some hope that low performing students’ trajectories in life can perhaps be effectively improved even after a difficult environment early in life.

Is the Chancellor right to relax pay regulation in the public sector?

Carol Propper

In a recent surprise announcement to the House of Commons the Chancellor announced that he wants to scrap national pay deals for public sector workers. Labour unions across the land are hitting back, arguing that this will damage public services. In fact, the evidence we have on the effect of national pay regulation suggests exactly the opposite.

National pay awards tend to overpay public sector workers in low cost areas of the country and underpay those in high cost areas. Recent research shows the size of these differentials. For example, the Institute for Fiscal Studies suggests that women working in the public sector in the West Midlands are paid upto 14  percent more than their private sector counterparts. What has received much less attention is that these pay differentials may have an important impact on the quality of public services provided in different parts of the country.

National pay arrangements effective impose a pay ceiling for workers in high cost areas. Simple economics suggests this should impact on the ability of the public sector to deliver services in these areas. The lower wages offered to public sector workers relative to their private sector counterparts in high cost areas will mean, all other things equal, that the public sector will struggle to recruit and retain the best quality workers. This in turn will mean problems in producing services.

Recent work undertaken at the CMPO and the London School of Economics confirms this intuition in a very stark setting. Analysis of the impact of national pay regulation of the wages of over half a million nurses in the NHS showed that hospitals in high wage areas had higher death rates for patients who were admitted following a heart attack.  Furthermore, the output of hospitals in low cost areas such as the North East did not appear to compensate for the lower quality output of their counterparts in the high cost South East. Our research suggested that deregulating pay to reduce the gap between nurses pay and that of their counterparts in the private sector would both save lives and cut costs. So in this case both economic intuition and the Chancellor’s instincts are right: deregulating public sector wages will improve the quality of public services.

Further details of the research can be found at: Propper, C and Van Reenen J (2010). Can Pay Regulation Kill? Panel Data Evidence on the Effects of Labour Markets on Hospital Performance. Journal of Political Economy 118 (2): 222-273.