Mar 19, 2018

Are the jobs in the centre of the city?

There’s a meme that most recent jobs growth in Australia’s largest cities is now in the city centre. Not true; a lot is because it includes the CBD, but nowhere near most

Alan Davies — Editor of The Urbanist

Alan Davies

Editor of The Urbanist

Jobs in Melbourne at the Census, city centre vs rest of metro area (source data: SGS)

Here’s Grattan Institute CEO, John Daley, claiming that “half of all jobs growth is now within a 2km radius of the city centres in both Melbourne and Sydney”. And here’s Professor Rodney Maddock from Victoria University telling us Melbourne has a serious problem because:

Most of its job growth is in the CBD, yet most of its population growth is at the far extremes of the city. It’s the same in other big Australian cities, like Sydney.

But there’re two problems with the assertion that “most”, or even “half”, of jobs growth is in the city centre:

  • First, it’s not true; most job growth is outside the city centre
  • Second, there’s much more to the geography of employment than “growth”; the great bulk of workers occupy existing jobs.

Let’s look at the Census data on employment for 2006 to 2011 in Greater Sydney and Greater Melbourne (GCCSA) and in their corresponding city centres (SA3).

It shows the municipality of Melbourne – the CBD and nearby areas – added 123,675 jobs over the ten years from 2006 to 2016. That’s a remarkable 42% increase over the period. It was certainly big enough to stress Melbourne’s CBD-focussed public transport system.

But what often gets overlooked by commentators is that jobs also grew in the rest of the metropolitan area. The number of jobs outside the city centre grew by a less spectacular, but nevertheless mighty, 30% over the ten years. Because it has many more jobs, the rest of the metro area added a whopping 381,325 jobs. That’s three times as many new jobs as were created in the city centre.

So, the city centre didn’t generate “most”, or even “half”, of Melbourne’s job growth over the last ten years. In fact, it accounted for 24% of all employment growth in the metropolitan area over this period.

The other important point is that looking solely at growth misses the main story. Only 20.5% of all jobs in Greater Melbourne were in the city centre in 2016. Notwithstanding faster growth, that was only marginally more than the 19.2% share it had ten years earlier. And it’s still a long way from the circa 24% of metro jobs the municipality of Melbourne had in 1981.

Put another way,  the great bulk of jobs in all of Australia’s capital cities are located outside the city centre. In 2016, 79.5% of jobs in Greater Melbourne were outside the municipality of Melbourne, down slightly from 80.8% ten years earlier. In fact, most jobs – around 70% – are more than 5 km from the city centre and half are more than 13 km from the centre.

The pattern was similar in Sydney between 2006 and 2016. Jobs grew 39% in the city centre compared to 25% in the rest of Greater Sydney, but again the increase in the number of jobs was smaller in the centre i.e. 138,562 vs 340,438.

In Brisbane and Perth – the next two largest cities – the city centre actually lost ground; jobs grew more slowly in both absolute and percentage terms than in the rest of the metro area. The share of all Greater Brisbane’s jobs in the centre declined from 20.8% in 2006 to 19.0% in 2016; the corresponding change in Perth was a large drop from 24.7% to 21.9%.

There’s no question the jobs concentrated at very high density in the city centre are extremely important for the regional and national economies. They tend to be high-pay, high-skill jobs. That’s the nature of the CBD. But it’s not where the great majority of jobs are located at present or are likely to be in the future. It’s not where the average worker will most likely find employment. We won’t have good policy on cities if we don’t get our understanding of the baseline right.

(Visited 88 times, 7 visits today)


Leave a comment

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

8 thoughts on “Are the jobs in the centre of the city?

  1. Horst (Oz) Kayak

    The statement; “We won’t have good policy on cities if we don’t get our understanding of the baseline right.” Is totally agreed with.

  2. Luke Elford

    Daley’s claim appears to relate to growth between 2006 and 2011. Part of the reason for the discrepancy in conclusions may be that there was a large increase in the ‘capital city undefined’ category between 2006 and 2011, which occurred for the various reasons explained here:


    A further complication, as Tony Morton notes, is the change in the ABS’s geographical classification system that occurred between these two censuses.

    The employment changes indicated in Daley’s graph (https://grattan.edu.au/wp-content/uploads/2017/01/Productivity-and-geography-Productivity-Commission-Dec-2016.pdf) for different distance bands sum to significantly less than the overall employment growth in Greater Sydney and Greater Melbourne indicated in the SGS analysis (about 105,000 versus 137,000 for Sydney; about 145,000 versus 206,000 for Melbourne). The discrepancy in each case is similar in magnitude to the increase in the number of workers in the ‘capital city undefined’ category (about 30,000 in Sydney and 51,000 in Melbourne).

    However, this cannot fully explain Daley’s claims about jobs growth in central Sydney and Melbourne.

    The increase in employment in Sydney LGA between 2006 and 2011, about 38,000, is still far less than the approximately 55,000 jobs Daley claims Sydney added within 2km of the CBD [1]. The corresponding increase in employment in Melbourne LGA, about 63,000, compares with a claimed increase in jobs of nearly 70,000 in central Melbourne [2].

    At any rate, comparisons of 2011 and 2016 data are more relevant to claims about current growth. There were no significant boundary changes between these censuses, but an imputation procedure has been introduced and the ‘capital city undefined’ category has been eliminated:

    http://www.abs.gov.au/ausstats/[email protected]/Lookup/by%20Subject/2900.0~2016~Main%20Features~POWP%20Place%20of%20Work~10069

    This means that comparisons of central SA3 and total greater city employment growth will if anything overestimate the percentage of employment growth occurring in the inner city. Even if one is very generous in including nearby SA3s, it’s clear that the 50% claim is outdated if indeed it was ever close to being accurate.

    [1] Including North Sydney LGA brings this up to 43,000.
    [2] Adding Port Phillip LGA reduces this to 60,000. The geographical area of Melbourne LGA increased slightly between 2006 and 2011, as Tony Morton notes.

  3. John Daley

    Circle with 2km radius is significantly larger than the SA3 for MELBOURNE or Sydney that SGS analyse.

    1. Alan Davies

      John Daley

      That might intuitively seem to be the case but it isn’t. A circle of 2 km radius has an area of 13 sq km. The City of Melbourne has an area of 36 sq km. Not surprisingly, there are fewer jobs within 2 km radius of Melbourne Town Hall than there are in the City of Melbourne.

      1. Tony Morton

        The tricky thing here is the SA3 ‘Melbourne City’ doesn’t quite coincide with the Melbourne LGA – it excludes the bit known as ‘Port Melbourne Industrial’. I’m also pretty sure there’s no working population data for this precise area in 2006, since the geography then was based on the old SSD/SLA structure. Another complication is that the 2006 Melbourne LGA excludes North Kensington, which was included under a subsequent boundary change.

        If I try to account for these factors as best I can in the ABS working population profiles, I find that the number of jobs in the Melbourne City SA3 increased by about 160,000 between 2006 and 2016. That’s still only about one-third of the total job growth in greater Melbourne over that time, rather than half, since the Melbourne GCCSA working population grew by about 500,000 over the same decade.

        Then there’s the curious question of how you measure the ‘number of jobs in Melbourne’. Do you count everyone who lists a workplace in greater Melbourne (regardless of where they live), or do you count everyone who lives in greater Melbourne and whose workplace status is ’employed’ (regardless of where they work)? The former seems a more natural choice, but it’s a curious question because the figure you get for this is _lower_ than what you get by counting total employed residents. Intuition would suggest the workforce size is greater than the employed resident population due to those who commute from regional areas, but it seems the reverse is the case. However, the ‘workforce size’ figure is certainly growing faster: as above, it increased by around 500,000 over the decade, whereas the employed resident population of the Melbourne GCCSA grew by only 400,000, and that of the Melbourne Urban Centre (the contiguous urban area) by only 300,000.

        Why should the workforce size be less than the employed resident population in Melbourne? I can’t say I have a good answer for this.

        Meanwhile, the Melbourne City SA3 is just one arbitrary definition of ‘central Melbourne’ that can be used for this exercise. As Alan has said, it doesn’t coincide with a circle of 2km radius, and arguably you should add parts of Port Phillip at least to get a reasonably uniform area in all directions. It’s nonetheless evident that central-city employment accounts for a sizeable part of the whole, whether or not it’s ‘half’.

        1. Alan Davies

          Tony Morton

          In terms of the argument being made here about ‘city centre’ vs ‘rest of Melbourne’, not a lot turns on whether the city centre is defined as Melbourne SA3 (31 sq km) or Melbourne LGA (37 sq km); same boundaries for the most part and both valid measures of the ‘city centre’. The key point is both are considerably larger than either a 2 km radius (13 sq km) or the CBD (8 sq km with Southbank and Docklands), yet they have nothing like “most” or “half” of jobs growth.

  4. Robert Harley

    Congratulations Alan. Good piece. Now what does that mean for alleviating traffic congestion?

Share this article with a friend

Just fill out the fields below and we'll send your friend a link to this article along with a message from you.

Your details

Your friend's details