Can estate renewal improve economic performance? It’s an important question given the tendency of policy makers to claim these economic benefits – even if the main aims of these projects are to improve the quality of housing, rather than to grow the local economy.
Unfortunately, our recent evidence review for the What Works Centre for Local Economic Growth suggested that the measurable economic impacts on local economies – in terms of employment, wages or deprivation – tend not to be large. Often, they are zero.
I was a little surprised, therefore, when the following headline popped up in my Twitter feed: “Government regeneration schemes add to productivity and growth. The Gorbals proves it”.
The piece in question, published earlier this week on CityMetric, summarises findings from a working paper by the Royal Town Planning Institute, which looked at the impact of the Gorbals renewal. Does this report, as the headline suggests, prove that such schemes add to productivity and growth? Unfortunately not. There are reasons to interpret these findings with extreme caution.
First, the study doesn’t actually measure productivity or growth. Instead, it looks at the percentage of the population that are income or employment deprived.
These are not measures of productivity. And while employment can contribute to growth, very local changes in employment as a result of area based schemes are hard to interpret (as we explain in our public realm briefing).
Second, the study doesn’t provide a before and after comparison. Instead, it looks at changes from 2000 onwards. This is problematic for a scheme starting in 1990, because we simply don’t know what happened in the first decade of the project. What if anyone that could – often the richest – moved away from the area to avoid the disruption that came with regeneration? In this case, changes from 2000 onwards would overestimate the impact of the scheme.
If, instead, the poorest families tended to be relocated early on, then the changes 2000 onwards will underestimate the impact. This is why we emphasise the importance of before and after changes.
Third, the study doesn’t use a convincing control group: an area suffering from similar problems, ideally both in terms of levels and changes, that for some reason happens not to get a renewal programme, while Gorbals does.
In fact, the study looks at one scheme and one not very well justified control area. This is a problem because, even if the scheme has no effect on outcomes, there’s still a 50 per cent chance it’ll do better than the comparison area.
This is the reason why impact evaluations look at multiple schemes and control areas. In fact, in this kind of “ex-post” study you can easily make the treated area look better simply by choosing a control area that has done worse over time.
The studies that we look at in our impact reviews avoid these problems. We considered over 1,000 studies claiming to measure economic impact, but found only 21 that meet our inclusion criteria.
These studies suggest that estate renewal programmes can increase property prices and rents – but they have limited impact in terms of improving income or employment.
What evidence is available is also pretty pessimistic on other outcomes. Evaluations that consider broader outcomes find that schemes tend to do little to reduce crime, raise levels of health and wellbeing, or improve educational outcomes among people living on estates.
Of course, that’s not to dispute the value of these projects in terms of improving the quality of housing and some aspects of the neighbourhoods that people live in. And with a larger body of proper evaluations, we might find convincing evidence of broader positive effects.
But the evidence that currently exists suggests that renewal schemes don’t boost the local economy. This doesn’t mean that we should stop doing estate renewal projects – but it does mean that we should be realistic about what such projects will achieve.
Professor Henry Overman is director of the What Works Centre for Local Economic Growth and a professor of geography at the London School of Economics.
This article is from the CityMetric archive: some formatting and images may not be present.