Executive Summary
The agglomeration benefits that can be achieved in dense employment areas, such as central business districts (CBDs), make cities an attractive location for many firms. However, the agglomeration effects produced in these areas may converge with a range of other favourable economic conditions. In this paper we examine the contribution of sectoral composition and workers' educational attainment to the observed productivity benefits that are currently being achieved in Auckland's CBD.
The major finding of this paper is that while sectoral composition and educational attainment would appear to contribute to the CBD's observed productivity advantages, employment density and the co-location of economic activity (or, in other words, "pure agglomeration benefits") are still the primary drivers.
Coupled with the modelling refinements that have been developed, this paper, building on our earlier research, has endeavoured to provide further insights into Auckland's productivity performance, and the nature of agglomeration benefits being achieved. The over-riding objective is to develop new perspectives concerning Auckland's role in central Government's economic transformation agenda.
Sectoral Composition and Productivity
The first step was to examine the impact of sectoral composition on productivity.
In 2001, the CBD contained just over a quarter of the jobs within Auckland City, with high activity in the "Finance and Insurance", "Communications Services", and "Property and Business Services", or what we term "Advanced Business Service" (ABS) sectors. This is balanced by the CBD's relatively low shares in the primary, manufacturing and wholesale, retail and health sectors however. The average earnings for all workers in the CBD is $47,100. This is approximately 15 per cent higher than the average for Auckland City ($41,000), and 27 per cent higher than the average for the Region ($37,000). With the exception of retailing, average earnings for all sectors in the CBD were higher than those in Auckland City, and the Region, as a whole.
The relatively high average earnings observed in the CBD stems from a combination of the area's particular sectoral composition (with a high proportion of high earning sectors), as well as the area's housing of many high earning activities within other sectors. In this regard, our analysis indicates that within sector productivity differentials account for approximately two-thirds of the difference between average earnings in the CBD and the Regional as a whole, with sectoral composition accounting for only a third. This implies that agglomeration benefits in the CBD are significant.
While there is a high concentration of ABS activity in the CBD, employment in other sectors is distributed relatively evenly. This may be evidence of a head-office effect, where activities co-locate because of a functional, rather than a sectoral, specialisation. In this regard, ABS activities could be functionally equivalent to the activities that constitute the remaining sectors in the CBD. Other centres with high levels of ABS employment (relative to other activities) include Newmarket, Ellerslie South and Takapuna, although the quantums are substantially lower than in the CBD. These centres have higher levels of concentration in more than one sector however.
Educational Attainment and Productivity
The second step was to examine the impact of educational attainment on productivity.
Here a clear linkage between earnings levels and levels of educational attainment has been identified. For almost all education levels, average earnings within the CBD are higher than average earnings within Auckland City, and the Region as a whole. Furthermore, the CBD workforce is typically better educated than that for the Region as a whole, both at an aggregate level, and for each of the sectors identified.
Education was estimated to account for about 15-20 per cent of the difference between productivity levels in the CBD, and the productivity levels in the Region as a whole.
Model Upgrades
Further to advancing the descriptive aspects of our research, we have also reviewed the models developed in our first paper. Three particular areas have been investigated: possible measures of accessibility, appropriate measures of activity intensity, and multiple linear regression analysis.
Back to Top