Where are all the jobs?

Economics
Where are all the jobs?
Keenan Jackson

Keenan Jackson

Urban Economics Consultant

Key Highlights

  • Comparing 2011 and 2016 Census employment figures can be misleading due to changes in data collection methods. In 2011, many workplaces couldn't be assigned to specific regions, whereas in 2016, the ABS used imputation techniques to allocate these, affecting direct comparisons.
  • Between 2011 and 2016, Australia's five largest metropolitan areas—Greater Sydney, Melbourne, Brisbane, Perth, and Adelaide—added over 500,000 jobs, with Sydney and Melbourne each experiencing around 10% growth.
  • While regional areas showed slower job growth, this aligns with their lower population growth rates. For instance, Greater Adelaide's job growth was only 1.4% during the same period.
  • For accurate employment analysis, it's advised to use modelled data from the National Institute of Economic and Industry Research (NIEIR), which accounts for methodological changes and provides a more consistent view of job trends. 

We hear a lot about the growth of jobs in Australia’s cities, but in this article, Keenan shows that’s only part of the story. This is a cautionary tale for anyone analysing how jobs growth in any part of Australia has changed over time. As always, experience and familiarity with the underlying data are critical to ensuring the whole story is told.

A closer look at employment data

When the second round of data from the 2016 Census was released recently, we were able to take a closer look at how employment trends have changed between 2011 and 2016.

Glenn, our Census expert, has already noted some preliminary findings from the data. However, in light of some recent articles about the growth of jobs in certain areas, I thought I would dig a bit deeper to explore some of these changes in greater detail (at smaller, more localised geographies), to counter some of the misstatements that have been made recently.

Some people have been reporting job change figures for regions by comparing the ABS Census numbers for 2016 with those reported for the 2011 Census. Unfortunately, this is a fundamentally flawed process that often results in misleading conclusions.

Why?

In 2011, a large number of people’s workplaces (621,676 to be exact) were unable to be coded to a Local Government Area (LGA) or region. In 2016, the ABS allocated those people based on an imputation methodology devised by the NSW Bureau of Transport Statistics, where locations are coded to a business based on associated data.

So, if you are simply comparing 2016 imputed data with 2011 non-imputed data you will get quite inaccurate results (probably leading you to think jobs in your region or LGA have exploded in the last five years!).

Considering this, and a number of other reasons for the Census undercounting job numbers, analysing change by comparing 2011 and 2016 Census jobs figures will likely yield inaccurate conclusions.

This is part of the reason we always recommend using Census data to explore worker characteristic trends (such as qualifications, income and method of travel to work), but NIEIR data, modelled using ABS Labour Force Survey data, for headline employment numbers.

Since it has been available, we have had 2011 (now 2016) Census data that was imputed to show jobs (by destination zone – a geography commonly used to analyse place-of-work data) freely available on our economic profiles under ‘employment locations’.

However, the imputed data does offer a more accurate picture of jobs changes between 2011 and 2016 at a regional or small area level because it’s comparing like with like (note: this comparison betweeen different time periods is not available on our sites, as inconsistencies in geographic boundaries make it likely that incorrect conclusions will be drawn).

So, what does it show?

Jobs change by region

As expected, jobs growth is being led by the large metropolitan areas with the 5 biggest capital cities adding just over half a million jobs.

However, there is a lot of variance between them. Jobs numbers in Greater Melbourne and Greater Sydney grew by around 10% in the five-year period. In contrast, Greater Adelaide grew by only 1.4%, below most regional areas.

While it appears that the non-metro areas are growing much more slowly, it should be acknowledged that the main working-age population (15-65 years) in metro areas grew by 9% between 2011 and 2016, compared to 3.1% in non-metro areas. So, a larger jobs growth should be expected.

ABS Census Imputed Jobs change by Region, 2011-2016

Region2011 Census imputed Jobs 2016 Census imputed Jobs 2011 – 2016 Change (Numbers) 2011 – 2016 % Change
Greater Sydney 1,998,983 2,209,296 210,313 10.5%
Greater Melbourne 1,868,6932,046,163 177,4709.5%
Greater Brisbane 980,222 1,037,901 57,679 5.9%
Greater Adelaide 552,433 7,883 1.4% 1.4%
Greater Perth 796,641850,158 53,517 6.7%
Total Big 5 Metros 6,196,9726,703,834506,862 8.2%
Rest of NSW 980,268 1,005,430 25,162 2.6%
Rest of Qld 949,056993,925 44,869 4.7%
Rest of SA 151,414 148,131 (3,283) -2.2%
Rest of Victoria 548,725 564,984 16,259 3.0%
Rest of WA 254,770 260,115 5,3452.1%
Total Big 5 States Non-Metro 2,884,233 2,972,585 88,352 3.1%
ACT210,695 220,645 9,9504.7%
Greater Hobart 97,564 99,361 1,797 1.8%
Rest of Tas. 106,688 106,813 1250.1%
Greater Darwin 60,887 72,120 11,233 18.4%
Rest of NT 37,773 33,860 (3,913) -10.4%
Other (territories and no fixed address) 411,014 474,621 63,607 15.5%
Total Australia 10,058,32910,683,842625, 513 6.2%

Source: Australian Bureau of Statistics, Census of Population and Housing 2011 and 2016.

Inner city agglomeration benefits

Looking more closely at the largest metropolises we can explore how inner-city jobs are growing (it seems to be a popular topic). You could use SA3 geographies. However, they are difficult to compare as the area they encompass for each city is quite different. For example, Brisbane Inner is only 13.5km2 compared to Perth City at a whopping 43.5km2.

So, to generate better comparisons, I have compiled SA2 level data. For Melbourne, Sydney and Brisbane it is necessary to combine some SA2s to get a roughly comparative geography. The dominance of the Sydney central city area is obvious, and the jobs growth in both central Sydney and Melbourne is far above that experienced in other cities.

In contrast, the inner areas of the other big cities appear to have hardly grown much at all or in the case of Perth even declined. The structural transition in the resources states appears to have impacted professional services jobs in their CBDs.

(It should be noted that while all areas contain parkland, Adelaide SA2 contains 5km2 or more of parkland and therefore actually has higher job densities in the built-up area than it might appear.)

ABS Census Imputed Jobs change by Inner-city areas, 2011-2016

Region2011 Census imputed Jobs 2016 Census imputed Jobs 2011 – 2016 Change (Numbers) 2011 – 2016 % Change
Sydney – Haymarket – The Rocks; North Sydney – Lavender Bay; Pyrmont – Ultimo; Surry Hills; Darlinghurst; Potts Point – Woolloomooloo (11.3km2)398,109463,46065,35116.4%
Melbourne; Docklands; South Bank; South Melbourne; Carlton (12.2km2)311,617 359,141 47,524 15.3%
Brisbane City; South Brisbane; Fortitude Valley; Spring Hill; Newstead – Bowen Hills (10.0km2)203,656 205,0441,388 0.7%
Perth City (10.9km2) 142,298 137,443 -4,855 -3.4%
Adelaide (10.5km2) 105,634 107,6121,9781.9%

Source: Australian Bureau of Statistics, Census of Population and Housing 2011 and 2016.

A good exercise would be to generate a definition of central cities based on combining Destination Zones (smaller than SA2s) and removing parkland/water from geographies (using ABS mesh block definitions). However, how do you define where central areas finish? Sydney’s business area effectively stretches across the harbour to North Sydney and beyond. Melbourne obviously includes Docklands and Southbank but also stretches north to Parkville and south down St Kilda road. Brisbane contains a number of high job density areas along the riverside in Milton.

Is it all about the central cities?

Clearly, not all inner-city areas are growing as strongly as others.

Looking at jobs growth across all SA2s in Australia, we can see again that inner Sydney and Melbourne created a large number of jobs during the intercensal period. However, out of the 25 SA2s with the largest jobs growth (number), only 8 could be classified as inner city or inner suburbs. Most of the jobs growth in Greater Perth and Greater Brisbane actually occurred outside the inner city.

While the jobs density may not be as intense in these other locations, it shows that other employment nodes such as Tech/Knowledge clusters (Macquarie Park, Parkville, Murdoch), Logistics/warehouse centres (Truganina), secondary metropolitan service centres (Parramatta, Chatswood, Dandenong), Regional cities (Geelong) and mining/resource areas (Ashburton, Weddell) all have a role to play. Many of them also don’t benefit from the scale of infrastructure investment (historical and current) that has gone into inner cities.

Top 25 SA2s for Imputed Jobs change (number), 2011-2016

Region (2016 SA2 Name)State Region Type 2011 Census imputed Jobs 2016 Census imputed Jobs 2011 – 2016 Change (Numbers) 2011 – 2016 % Change
Sydney – Haymarket – The Rocks NSWInner City 270,223 320,829 50,606 19%
Docklands VictoriaInner City 34,858 57,555 24,16072%
Melbourne VictoriaInner City 201,734 221,136 19,402 10%
Ashburton (WA) WA Remote12,055 23,540 11,485 95%
Macquarie Park – Marsfield NSW Middle Suburb 41,481 48,394 6,913 17%
WeddellNTFringe 8987,4846,586733%
Murdoch – Kardinya WA Middle Suburb 7,632 13,618 5,98678%
DandenongVictoriaOuter Suburb 59,901 65,6885,78710%
ParkvilleVictoriaInner City 23,049 28,207 5,15822%
Pyrmont – Ultimo NSWInner City 31,97436,8314,85715%
North Lakes – Mango Hill QldFringe & Peri-Urban 5,553 10,3114,758 86%
Baulkham Hills (West) – Bella Vista NSWOuter Suburb 18,912 23,584 4,672 24%
Surry Hills NSWInner City 29,818 29,818 4,654 18%
TruganinaVictoriaFringe & Peri-Urban 2,817 7,207 4,390 156%
Richmond (Vic.) Victoria Inner Suburb 32,47236,837 4,365 13%
Ormeau – Yatala Qld Fringe/Outer Suburb 14,342 18,438 4,096 29%
Geelong VictoriaRegional City 27,306 31,099 3,778 14%
Chatswood (East) – Artarmon NSWMiddle Suburb 23,93127,655 3,724 16%
Newstead – Bowen Hills Qld Inner City 16,643 20,347 3,704 22%
Madeley – Darch – Landsdale WAOuter Suburb 14,76418,357 3,593 24%
Parramatta – Rosehill NSWMiddle Suburb 46,819 50,227 3,408 7%
Deer Park – Derrimut VictoriaFringe/Outer Suburb 9,627 12,934 3,30734%
Concord West – North Strathfield NSWMiddle Suburb 16,587 19,823 3,236 20%
Rockbank – Mount Cottrell VictoriaFringe & Peri-Urban 2,6165,740 3,124 119%
ClaytonVictoriaMiddle Suburb 31,244 34,2032,9599%

Source: Australian Bureau of Statistics, Census of Population and Housing 2011 and 2016. ABS Census 2011 Destination Zone job counts have been matched to 2016 SA2 boundaries. Some job allocation has been applied to best fit Destination Zone boundary changes between 2011 and 2016.

If you’re working in Docklands, Ashburton, North Lakes or especially if you’re in Weddell in the Northern Territory, we would love to hear from you! How has the massive growth in jobs in your area affected you and the people living there?

If you would like to explore economic or demographic issues in your area or broader region, contact one of .id’s consultants or learn more about how we can help you put the story of your area together.

.id is a team of demographers, urban economists, spatial planners and data experts who use a unique combination of online tools and consulting to help governments and organisations understand their local economies. Access our free economic resources to help profile your local economy.

 

STAY INFORMED

Subscribe to monthly updates