Academic journal article Australasian Journal of Regional Studies

Spatial Effects of 'Mill' Closures: Does Distance Matter?

Academic journal article Australasian Journal of Regional Studies

Spatial Effects of 'Mill' Closures: Does Distance Matter?

Article excerpt


Any small town with employment heavily concentrated at one employer (the 'mill') has the threat of mill closure as an ever-present concern. We examine the impacts of two large mill closures in New Zealand that occurred in the early-mid 1980s in towns with some contrasting features. Using a difference-in-difference approach, we examine how the dynamic adjustments differed across the two towns and relate these contrasting adjustments to features of the towns and their populations.

Our purpose in this analysis is twofold. First, understanding factors that affect adjustments to negative employment shocks is important in its own right, especially where policymakers aim to facilitate improved economic and social responses in the wake of such a shock. Second, the impact of a major exogenous closure can tell us a considerable amount about the benefits of a similar mill opening. Where the closure is unanticipated (as in our examples), we can treat the observed impacts as akin to the outcomes of a natural experiment, whereas this is more difficult in the case of an opening where many other (positive) factors may be operating in tandem with the mill opening.

Drawing on a range of prior research, we derive a set of hypotheses concerning adjustment dynamics of employment and population, and confront these hypotheses with data from the two meat works plant closures at Patea and Whakatu. We analyse the effects of each closure on population, housing and employment within each town. In each case, we express the town data relative to (two sets of) comparator data, detrend the resulting ratio, and then examine the temporal differences in the responses of the detrended ratios. This difference-in-differences approach enables us to isolate the adjustment components due to the respective mill closure. The results assist in understanding the role of major investments in rural areas, and provide insights for regional development policy makers who may be considering the location of such investments.

We provide a brief survey of related studies, together with descriptions of the two closures and hypotheses, in the next section. Section 3 outlines our data and methodology. Results are presented in section 4, with conclusions following.


In New Zealand and internationally, regional development programmes have often attempted to promote the construction of large processing facilities (mills) in areas away from major urban areas. The rationale for their establishment in non-urban areas relates to the net economic and social benefits these developments may bring to rural towns. These benefits include employment, higher revenues, urban growth, and better education and health services. In effect, the mill is a form of infrastructure, often servicing multiple suppliers (as in the case of meat works or dairy factories) and providing the commercial substance to attract other service providers to the locality. In this latter respect, they are similar to a transport link that attracts new firms and population to an area. However, the effects associated with the closure of these facilities on the towns in which they are located can be devastating. Given the immobility of houses and commercial properties, these structures are not likely to disappear as quickly as they emerged; many may fall into disrepair and cheap rents may encourage those without employment to move in.

Prior Literature

Our analysis of adjustment dynamics following mill closures is conceptually related to analyses of the impacts of regional employment shocks. Blanchard and Katz (BK, 1992) analysed regional adjustments to employment shocks using United States state-level data. They found that the dominant adjustment mechanism following an employment shock is labour mobility. Their study established that US employment shocks have a permanent component to them. In the case of a negative employment shock, the employment response after 5-7 years is almost entirely reflected in net outward migration from the region, leaving the unemployment rate and participation rate relatively unaffected. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.