1. Introduction
The primary purpose of this paper is to provide a preliminary examination of the agglomeration effects taking place in Auckland, by outlining a range of simple relationships. While the paper has been developed with a policy audience in mind, it would likely be of interest to a wider audience given the intent is to promote a better understanding of Auckland's role in achieving Government's economic transformation agenda. As with the first phase work, further economic analyses would be necessary to refine and develop our findings.
1.1 Background
This study builds on our earlier, Phase 1 paper produced for the Ministry of Economic Development (MED) entitled: Assessing Agglomeration Impacts in Auckland. Presented to the MED in early 2007, this work covered both a review of the theoretical literature and a review of the empirical evidence for agglomeration in Auckland.
The main conclusions from the Phase 1 research were:
- The analysis of the position in Auckland suggests that there are strong linkages between accessibility and density, and density and productivity (using average incomes as a proxy for labour productivity), in line with findings in the international literature.
- For Auckland City, the relationships (between accessibility and density, and density and productivity) are particularly strong and relatively robust. The CBD combines high employment densities and high incomes, while similar conditions, outside the central area, are evident in Ellerslie South.
- The relationships in the other cities in the Auckland region are less strong, but still show a positive relationship between density and productivity.
- A similar pattern emerges when looking at the relationship between accessibility and employment density, with a reasonably strong relationship in Auckland City, and a weaker but still positive relationship elsewhere.
1.2 Focus Areas
The results reported in the Phase 1 study considered employment as a whole. It did not consider, in any detail, the make up of employment and the way in which this might impact on agglomeration. In this Phase 2 research, employment characteristics, taking into account the sectoral makeup of employment and the educational levels of the workforce are considered, along with the extent to which these factors appear to contribute to productivity differentials.
Within the Auckland Region in 2001, the base year for this analysis, Auckland City provided approximately 46% of total regional employment. As our earlier analysis indicated, Auckland City displayed fairly robust relationships between accessibility and employment density, and employment density and productivity. BeCAUse of this, we have concentrated on the detailed position for Auckland City, and, in particular, the position for the CBD. However we have also taken the opportunity to briefly look at the position for selected centres in the other cities within the Region.
This analysis is based on information by workplace, derived from the 2001 census.
Appendix A lists the categorical breakdowns for both our sectoral and educational analyses.
1.3 Methodology
1.3.1 Method
The major assumption in our methodology is that:
wage = marginal product of labour.
This follows the intuitive logic that over the long-run workers will be appropriately compensated for the value of the output they produce. Thus productivity is proxied for by wage levels. This is a common approach in the literature, and was used by Lewis & Stillman in their assessment of regional economic performance for the Treasury.1
The identification and quantification of the relationships set out in Figure 1.1 provides a means of assessing the potential effects of changes in accessibility on the potential employment densities and on the consequent levels of productivity and output. Where the accessibility to existing major centres is improved, the impact may be substantial. Currently the model used to assess these relationships does not distinguish between different type of jobs or different education levels, although possible refinements could include these factors.
Figure 1.1: Forecasting Model
Using the quantified relationships, the potential effects of changes in accessibility on employment density can be assessed, with improved accessibility giving rise to forecast changes in the numbers of jobs likely to be achievable in specific areas. This increased employment will generate the increased output from the additional workers, and by raising employment densities, will also contribute to an increase in productivity and output for the existing workforce. Both these factors will contribute to an overall increase in economic output for the areas affected.
The increase in total output estimated using this approach is not captured by existing conventional appraisal methods, particularly as is typically the case when fixed land uses are assuMED. In principle, therefore, these benefits represent an addition to the total economic benefits of a transport project as conventionally calculated.
1.3.2 Causality Issues
Whether density leads to productivity, or productivity leads to density, is continually debated. However, the existing literature gives support for focusing on the former.
The empirical literature on the foundations of agglomeration, has only been able to correlate increasing densities with rising productivity. However, the theoretical literature gives support for the density-productivity relationship. In this way the externalities of knowledge spillovers, labour markets, and input-output linkages, enabled by matching, sharing or learning mechanisms, are seen to arise in dense employment areas, leading to increased productivity. Graham has examined the density-productivity relationship in the UK, but, due to data limitations, cannot rule out the possibility of reverse causality. However, Graham recalls the research of Ciccone & Hall, and Ciccone, who in their tests for endogeneity (where density increases as a response to productivity increases) found no "substantial evidence" of it occurring.2
As a part of the continuing MED-Ascari work programme, we have identified a research process that may help to provide some clarity on the causality debate. In this respect, we propose primary research be undertaken with business owners in Auckland's CBD to assess the degree to which their location decision was influenced by the advantages that stem from co-location with other businesses, as well as time series research which will provide a clearer impression of the relationship between increasing densities and rising productivity. This is further outlined in section 5.
It should be noted however, that this is a cross-sectional, associational study and, as a result, reverse causality cannot be ruled out.
1.3.3 Measures of Employment Intensity
To date, our analysis has deployed employment density as the appropriate measure of employment intensity. In this respect, our approach examines the relationships between accessibility and employment density, then the relationship between employment density and productivity. However, it has been suggested that employment scale may be a more suitable measure of employment intensity, and provide a better relationship with productivity. In the course of this phase of the research, we have taken the opportunity to use the empirical evidence for Auckland to explore this issue.
1.3.4 Defining the CBD
A key focus in the work has been understanding the economic role and effects of the CBD, including the extent to which its sectoral composition differs from that of the rest of Auckland City or the Auckland Region as a whole.
For the analysis of the position within the CBD, we have broadly followed the standard definition of the CBD as the area bounded by SH16 and SH1 (as developed by Auckland City). However, the analysis we have undertaken is based on Census Area Units (CAUs), which, in some cases, straddle the motorway. For the purpose of this work, the CBD has been defined to include the Census Area Units of Auckland Harbourside, Auckland Central West, Auckland Central East, Newton and Grafton West. The area included is set out in Figure 1.2.
Figure 1.2: Definition of the Auckland CBD
→ Full size version of Figure 1.2 [226 kB JPG]
1.4 Order of the Paper
This paper has two over-riding themes. Sections 2 and 3 provide a descriptive account through an examination of the nature of agglomeration effects in Auckland. Chapter 4 then focuses on modelling and analytical issues.
Back to Top