Time for Tables

My final AFR oped for the year is on school reporting (aka league tables). Full text over the fold.

ABC Brisbane called me this morning to do an interview on the piece. At the time, I was at home looking after a rather tired 10 month old, who was exhausted from crawling around, and merely wanted to be held on my hip. So what resulted was a radio interview with me trying to sound serious about school productivity, and Sebastian interjecting “ohh” and “eh” at inappropriate moments. Fortunately the interviewer took it in good humour.

Schools need a report card too, Australian Financial Review, 20 December 2007

As Prime Minister Kevin Rudd sits down with premiers and chief ministers at the Council of Australian Governments meeting in Melbourne today, one of the topics of conversation will doubtless be how to implement Labor’s “education revolution”.

Past education policy debates in Australia have often been unproductive. While the Howard Government proposed some useful reforms, they were ultimately unable to find a way to work with teacher unions, and found themselves caught up in distracting debates about Maoism. For its part, Labor’s rhetoric can give the impression that the party believes reform requires little more than opening up the funding spigot.

Breaking the ideological deadlock requires attention to the new productivity agenda in Australia: making public services work better. In the case of schools, there is strong evidence that reform is needed. Despite a significant increase in funding, literacy and numeracy scores of Australian teenagers have failed to rise over recent decades. On average, new teachers are less academically talented today than they were two decades ago.

Boosting the performance of Australian schools is far from straightforward, but one sensible reform would be to begin reporting on the performance of individual schools, so that parents can better choose between their local schools. Such a reform would bring us into line with Britain and the United States, where policymakers across the board take the view that a school’s test scores are quintessentially public information.

Now, new research has shown that better information has direct benefits for children. In a novel experiment in North Carolina, Yale University economists Justine Hastings and Jeffrey Weinstein randomly provided some parents with more information about the quality of their local schools. They found that that one in twenty parents responded to the additional information by switching their child into a better school. Notably, African-American parents were more responsive to test score information than white parents (perhaps because they had less school information to begin with).

But does school choice really benefit kids? The following year, Hastings and Weinstein followed up the children who switched schools, and compared their test scores to the non-switchers. They found that moving to a better school raised test scores substantially. In other words, the new schools didn’t just skim the cream; they added value. 

In Australia, very little information about school performance is presently available. Some states only release information on the top students, while others provide data to newspapers on the condition that schools be listed alphabetically. Most public information relates to year twelve, though the Western Australian government publishes a website with primary school results shown in a graphical format. The most restrictive rules apply in New South Wales, where an infamous 1997 Daily Telegraph headline (“The class we failed”) has stymied test score reporting for over a decade.

Fortunately, there now seems to be a bipartisan federal consensus for change. Prior to the election, Labor’s then education spokesman Stephen Smith said that a Rudd Government would attempt to make available test score data at a school level for literacy and numeracy tests in grades 3, 5, 7 and 9. Over the next few months, federal Labor can expect pushback from the states and territories, and should have its answers ready.

Some critics will argue that test scores aren’t all that matter. True, there is more to education than standardised tests, but a thorough knowledge of the basics complements critical reasoning. Moreover, plenty of research shows that employers prefer to hire literate and numerate workers. In the North Carolina experiments, Hastings and Weinstein estimate that switching to a better school may end up raising students’ lifetime incomes by as much as $100,000.

Others will claim that raw test scores don’t provide useful information. The simple answer to this critique is to produce what Bill Louden of the University of Western Australia calls “smart” league tables, which are adjusted to account for socio-economic status, or which measure value-added.

Another common criticism is that parents can always visit the school to get more information. Yet for disadvantaged parents, making an appointment with the school principal can be daunting. As the North Carolina experiments suggest, making test score data readily available may well benefit the underprivileged most of all.

As a first stage in the education revolution, Kevin Rudd and Julia Gillard should bring more sunlight into the schooling system, making public all test score data for all schools in Australia. School league tables are no magic bullet, but you can’t have a revolution without information. 

Dr Andrew Leigh is an economist in the Research School of Social Sciences at the Australian National University.

Advertisements
This entry was posted in Economics of Education. Bookmark the permalink.

6 Responses to Time for Tables

  1. conrad says:

    I imagine that whilst releasing school scores might help the system a little overall, versus just moving kids in circles (perhaps parents will be willing to pay a little more to send their kid from one school or another), it seems to me that it is really a tiny issue compared to the one sitting on the bottom of the third paragraph “On average, new teachers are less academically talented today than they were two decades ago”. (Which I might say was a very polite way of saying it cf. “Our median primary school teacher scored worse than 80% of other school leavers, including those that failed i.e., is probably quite dull” — or whatever the real number is this year).
    I might also point out that league tables have been an overall negative for universities, as they have caused extra administrivia and poor allocation of resources such that universities can get up these tables. It will be interesting to see the extent that schools are willing to do this type of thing as well. If they really start burning cash to get up the tables, it is possible to imagine an overall negative effect.

  2. JM says:

    “They found that that one in twenty parents responded to the additional information by switching their child into a better school.”

    Andrew, I wonder if the word ‘better’ isn’t a bit of a fallacy here. I can’t find it just now, but there was a wide study done in the UK a couple of years ago that came to the following conclustions:-

    * the measured differences between many schools (particularly those in the same area) are very slight and well within the margin of error,

    * those small differences show random variations from year to year,

    * when parents (particularly of better students) started switching they amplified the previously non-existent differences so they became significant (for the usual reasons: more attentive students, better motivated parents who put more time and effort into the school)

    * since some school funding was linked to performance, the ‘better’ schools also became better resourced further amplifying the effect

    In other words, league tables – far from improving information and the efficency of the market – tended to distort the market based on initially purely random effects.

    Can this really be a good basis for public policy when governments have a responsibility to all students?

  3. Pingback: CoreEcon » Blog Archive » Winding down the year

  4. Kevin Cox says:

    This would be a good start Andrew.
    The important feature is giving parents the ability to make an informed choice. We can build a better “market” if schools are able to choose or reject students and are rewarded for taking on children whom no one else wants and are rewarded for increases in performances of children while at their schools.

    A school can extend the measures of performance to “performance of parents” in supporting the school and in trying to measure other social aspects. (Perhaps hugs per day:) to help parents make an informed choice.

    My guess is that it is not necessary to force schools to publish measures. They will do it themselves if they see benefit in having waiting lists so they can choose students. What is important is to have independent assessments so that choice is informed. That is, if a government made independent measurement available for no cost to schools with schools allowed to decide themselves how to use the results then the rest will start to take care of itself provided both parents and schools have a genuine choice.

  5. conrad says:

    Kevin,

    you can give parents as much choice as they want, and no doubt its good at that level, but what are you going to do with kids that no schools wants? I doubt any school would want disruptive kids, even if you offered them twice the funding.

    Here is the big problem as far as I see it:
    First, to simplify things, lets say each kid needs 1 teacher.
    I am now the government, and I give you 4 kids but only 3.5 teachers. As you can see, we are down .5 of a teacher, and there arn’t actually any others to be found, and nor would you be able to get any for the money you have anyway. So you can either try and make the other ones more efficient (probably unlikely, because 2 of them have IQs of 87, and only became teachers because thats the only thing they could get into at uni), or get more money in the long term so that teaching becomes more attractive and you have enough teachers.

    I don’t see how this problem solves itself out by allowing more choice (which is not to say I’m against more choice).

Comments are closed.