The CofFEE/URP Employment Vulnerability Index – with updates

Today I released a major new research report Red alert suburbs: An employment vulnerability index for Australia’s major urban regions which was the result of a collaboration with Scott Baum (URP, Griffith University) who I share a large ARC Discovery Grant with. The Report and its findings has already received front page coverage in the large Australian dailies – The Age and the Sydney Morning Herald.

The CofFEE/URP Employment Vulnerability Index (EVI) identifies the suburbs across Australia that are most vulnerable to job losses as a result of the current economic crisis.

We developed an economic model to predict the employment vulnerability suburb by suburb. The EVI covers all the Capitals and major regional cities – that is, around 70 per cent of the total population.

The EVI reveals those suburbs – Red Alert and Amber Alert suburbs – which are most exposed to potential job losses and least well placed to escape disadvantage associated with increasing unemployment. The high risk suburbs include the traditional battler suburbs. But what is disturbing is that we have also identified a new arena of socio-economic disadvantage that will emerge as a result of the current crisis. The new arenas include the mortgage belt suburbs that have grown on the periphery of our major cities. So the neo-liberal debt binge will have a huge hangover in these types of suburbs if the economy goes into recession.

We also identify those suburbs that have Medium low risk and, in turn, Low risk of job loss. We have created detailed maps to show the spatial patterns of these four job-loss risk categories by the cities we examine (Capital cities plus regional centres with more than 20,000 people).

The Report is accompanied by a fully-resourced WWW Home Page which has extensive maps; tables you can construct to suit your regional interest; and detailed data profiles of the job loss suburbs you can select for viewing. The Final Research Report with additional data can also be downloaded from the EVI Home Page. Further developments in the near future will include interactive maps which will provide a more dynamic understanding of urban disadvantage in Australia’s metropolitan and regional centres.

Feel free to visit the Employment Vulnerability Index Home Page for maps and tables and the Final Research Report.

All comments will be appreciated.


I had a very long day answering media enquiries – around 50 odd interviews from Esperance to Hobart to Tamworth and many other places in between. Some of the comments I have seen on the ABC national news page (see link below) are critical which is good but miss the point of the exercise. See my comments to Lefty for example about Fyshwick which seems to have upset some Canberrans.

In the course of the day I have also learned that some areas are pursuing special infrastructure projects (for example, Shepparton is embarking on a $2 billion pipeline project) which will attenuate the job risk exposure that our “broad brush” analysis generates. The data and maps should always be interpreted with the greater insight that localised knowledge provides. We do not claim to be experts on every suburb. But in terms of historical relationships – that is, factors that drive job loss – the analysis we have provided will be broadly reasonable notwithstanding these local factors that might modify, negate, amplify our assessments.

No empirical work is without simplifying assumptions. It is important that end-users are aware of the strengths and weaknesses of the research being presented to them as “science” and use their own judgements accordingly.

Some media links from today

Update 2 – Monday evening

The ABC Adelaide ran this story on our work under the heading New list of suburbs at risk, which I think should be emphasised – at risk!

I was quoted by the ABC as saying:

Out to the north-east, you know, Yatala Vale and Para Hills West. Suburbs to the south like Craigmore and Reynella and Woodcroft, these are what we call areas of emerging disadvantage.

They’re likely to become battler suburbs in the future unless there’s major policy intervention.

These are areas that we haven’t traditionally considered to be disadvantaged but they are now equivalent, in terms of risk of very high job loss, to the traditional battler suburbs.

Note again – high risk of job loss rather than a prediction that there would be high job losses.

One commentator from Adelaide University John Spoehr criticises the report for “failing to take into account the delay in an expected mining boom for South Australia.” Spoehr apparently said that he expected “unemployment to rise more slowly in SA than in other states, where a mining bust has led to job losses.”

To which I say – have you read the Report John? If you have you haven’t understood it and if you haven’t then you should not comment on our work. We make no predictions of job loss. We merely indicate where there is risk of job loss, other things equal. Again if there is local knowledge about offsetting factors, or the recession is mild or whatever, then there may be very few job losses. But the risk remains that Red Alert suburbs are more vulnerable to job losses if the economy enters a protracted recession.

Then the politicians decided to enter the fray. The South Australian Employment Minister Michael O’Brien claims the “the data used in the study is old and the report overstates the likely severity of the downturn.”

Wrong and wrong.

First, the data is the 2006 Census of Population and Housing, the latest available the Suburb level. The analysis assembles known risk factors into a composite index using standard and accepted statistical methodology. It is highly unlikely that there have been significant shifts in the main drivers of job loss risk in the period since the Census. These factors tend to evolve relatively slowly. So the age of the data is rather irrelevant. I wonder if his Department is ever using the latest Census data for anything? I hope they are but if Michael was correct he should not be basing a single policy development on it. His Department will be and so his comments are just cant.

To give examples: The percentage of those without post secondary schooling is a high risk factor – do you think that has shifted much spatially since August 2006? Do you think the composition of industry in a region has shifted dramatically since August 2006? Do you think there have been that many gender reassignments in regional Australia since August 2006? Clearly, some things change quickly and render Census data inapplicable. But for the purposes of this study which is to get a suburb-by-suburb snapshot of vulnerability to job loss, the accusation that the 2006 Census data is too old is wan!

Second, we make no predictions about the severity of the downturn or the magnitude of any possible job loss.

Employment Minister O’Brien then was quoted as saying:

The way they’re talking is a situation that’s going to be in place for the better part of a decade. There’ll be some short-term pain, that’s undeniable, but to talk about whole suburbs around Australia slipping into disadvantage I think is sensationalist …

Well, recessions do have aftermaths that easily last for a decade in areas that are most affected. The national unemployment rate took more than 14 years to get back to the pre 1991 recession level. Underemployment has never returned to its pre-1991 level (at the top of the boom it was approximately 4-5 per cent of the labour force yet was around 1.5 per cent (maximum) before the 1991 recession. There is huge short-term pain associated with a recession for sure. But then for many disadvantaged groups the pain persists for years. If a suburb is really affected by high unemployment then it is not alarmist or sensationalist at all to assume these negative effects will be long-lasting and intergenerational.

The last point is worth emphasising. Other research I have done shows that a young person who grows up in a jobless household has much higher risk of being disadvantaged as they go into adult life. The children of the 1991 recession unemployed show up in, for example, HILDA data, as just that.

Comments policy

By the way, I will post all comments that add to the debate. I will delete all comments that are just personally abusive.

This Post Has 8 Comments

  1. Hi Bill

    I haven’t actually read your report yet, but do you think that employment vulnerability in some places may be compounded by some measure of “downturn-driven urban drift”? Resource boom areas such as mine (Gladstone – central QLD coast) typically see many people heading off down to Brisbane as employment opportunities dry up and I expect that all the regional areas experience a similar phenomonen in a serious or protracted downturn.

    An influx of job-seekers from outside wouldn’t be very helpfull to those in the red and amber zones. Many of those from areas such as mine will be blue collar workers, formerly employed in mining, construction and industrial plant processes – and accustomed to recieving reasonably hefty pay packets.


  2. Hi Bill,
    what some researchers might not know is that the majority of population in these red zones are actually full time employed for cash. they just happen to be on the dole at the same time.

  3. Dear Lefty

    The Employment Vulnerability Index is a static measure – which is derived from a weighting of known high risk factors. It just tells us that a region has a particular risk. It might not have many people living in it (for example, Fyshwick in the ACT is classified Red Alert – we know it is largely an industrial area some while there will be significant relative job losses there the actual number may be small). So the sort of population movements that you describe are not taken into account by us.

    I have done other work tracking for example who benefited from the employment growth that has now come to an end by modelling changing commuting and migration patterns. We know that, for example, a number of low skilled jobs created in they Sydney area between say 1996 and 2004 went to commuters and movers coming in from say Newcastle, Central Coast and Wollongong who were escaping poor labour markets in their home regions. The longer commuting was done by higher skill workers who could not find work in the more depressed regional areas (like Newcastle) and were able to out-compete the lower skilled workers in Sydney and take jobs that would normally be considered low skill work. We call this process “bumping down” – the more skilled bump the less skills down the pecking order and take jobs that the low skill would normally have.

    So unless there is adequate jobs growth spatially distributed to match the social settlement (where people live) then you get these perverse population shifts that entrench disadvantage in the out-commute/migrate region and stop the low-skilled from benefiting from the growth in their local regions.

    best wishes

  4. Thanks Bill.

    What do you think might be the motivation behind this “bumping down” effect? Employers being able to employ a more highly-skilled person at low cost but then pressuring them to perform more highly skilled tasks that lie outside the job description of the worker they have replaced? Employers seeing the opportunity to get “more bang for their buck”?

  5. Dear Lefty

    I am not sure about that. I have actually researched that question. We have identified the phenomena in our work but trying to get to the motivation might be quite difficult. But it is an interesting question for sure. A simple explanation, and often the most simple reasoning is the best, is that they think a higher skilled person will be more motivated and more disciplined. But then you often hear of the anecdotal claim “s/he is too qualified for the job”. So it is unclear. Perhaps your hypothesis is worth considering. There is no doubt that the trend has been to pressure workers into increasing volumes of unpaid work. That would be consistent with your conjecture. Maybe I will work out a way to research that question one day.

    best wishes

  6. Cheers Bill.

    Yeah, I didn’t mean to suggest that it was the only motivation, just one of a number of possibilities.

  7. BTW, I have posted a question in “Defecit spending 101 – part 1”


Leave a Reply

Your email address will not be published. Required fields are marked *

Back To Top