Seeing trees and forests: a comparative evaluation of business clusters and national industry associations in the New Zealand forest sector.

This discussion suggests a number of attributes that potentially differentiate how enterprise managers perceive the relative effectiveness of national and local associations:

* Clusters may be seen as providing more opportunity for interaction with other business members than are national associations. This possibility arises from a cluster's localised membership, which may mean that other members are known and that active participation is less demanding than with a national group.

* The ability to influence group activity may be higher in the case of a cluster than national association. This possibility follows from the ability to maintain participation. It may also result from it being easier to obtain agreement about group activity in the case of local initiatives than a national group which potentially has a wide range of issues to address.

* An industry association may be better placed than a cluster to develop collective resources such as industry standards, marketing campaigns and industry development strategies. Non members are hard to exclude from the benefits of collective service and this is a disincentive for their production. Given that clusters may involve only part of the national industry, they potentially have the greater disincentive to invest in collective services.

* Industry associations may be viewed as the more effective vehicle for informing managers about industry regulation and support programmes. This follows from the national focus of an industry association. Conversely, clusters may be perceived as more effective in engaging with local government.

* Industry associations may be perceived as political organisations whereas business clusters are viewed as more focussed on business development rather than industry politics. This would follow from an industrial association's involvement in lobbying activity, their possible capture by particularly motivated or well resourced members and from having a longer history than most cluster groups.

* Membership of a cluster may be seen as implying more commitment to the group than industry association membership. National association membership may involve little more than an annual subscription where as the pressure to 'join in' may be strong with a locally based cluster group.

CASE STUDY SELECTION AND RESEARCH DESIGN

The context for this study was the particular history of national industry associations and cluster groups in New Zealand. National associations have been a longstanding feature of the business environment whereas cluster groups are a recent product of public policy support (Perry, 2001; 2004). The diffusion of cluster groups has varied between sectors. For the purpose of this investigation, an industry with a number of cluster groups was looked for to obtain a sufficient sample of enterprises with exposure to a cluster group. The forestry sector met this criterion as it gave rise to five projects that gained recognition under a government cluster development programme, the four largest of which were covered in the study (Table 2). In addition, it is also a sector with a range of national associations (Table 3).

Association with a cluster group was the starting point for the investigation. It was judged that all enterprise managers in the forestry sector would be able to offer informed comment on the perceived effectiveness of an industry association. Similar comment on a cluster group was seen to be more dependent on direct exposure. As well, the prioritisation reflected the lack of a formal membership structure for most cluster groups. Typically, a public agency had sought to promote a cluster group on the basis of building an inclusive association using their resources to limit dependency on membership fees or other member contributions. This made it possible to identify a range of enterprises with a varying degree attachment to a cluster. Had the study commenced with a sample of industry association members, there would have been more restriction of the sample to a group of enterprises with a demonstrated commitment to at least one of the forms of association.

Using the prior understanding of the potential roles and challenges for national and local groups, as discussed in the previous section, a set of comments were devised that identified potential strengths or weaknesses of each form of association. Individual enterprises managers were interviewed and asked to indicate whether they agreed or disagreed with the statement. In addition, respondents were questioned about their level of support for the two types of group and, where relevant, asked to identify actual outcomes from membership. The study covered four of the five cluster groups existing in the timber industry, excluding the smallest group known as Forestry Wairarapa. This group comprised seven forest management companies including local councils that have small forest holdings maintained for reasons of land conservation rather than as a commercial business.

The use of face-to-face interviews to collect data reflected a number of considerations. Practically, cluster-linked businesses are located in relatively close proximity to each other enabling a single researcher to complete a comparatively large number of interviews over a short space of time. Given the possibility, personal interviews were preferred to gain the cooperation of potential respondents. Particularly with regard to the evaluation of cluster groups, respondent sensitivity to providing evaluative judgements was a potential threat to the validity of responses. Use of a qualitative method was primarily to add confidence in the quantitative data rather than to investigate individual experiences in depth. Interviews were used to explain the questions in the survey and check that responses were based on the intended meaning of the question. A feature of the study was the high level of triangulation possible between the responses given by members of each cluster group. Individual respondents frequently made reference to their understanding of the views of others in their cluster and commented upon each other's actions. At the same time, as enterprise managers were generally well known to each other reporting individual responses is constrained by the need to maintain confidentiality. This partly explains the preference to rely on the quantitative scores obtained rather than reporting qualitative evidence in the form of individual quotations.

A complete survey of organisations linked to the clusters was not attempted. The preference for face-to-face interviews and need to fix meetings over a limited time was a practical constraint on completing a census. As well, it was not considered relevant to cover all non business organisations (such as training agencies, port managers and local authorities) as these participated as 'associate members' and most would not have experience of a national forest industry association. The priority was to interview businesses that had been identified as a supporter of the cluster group. Of 56 organisations approached for an interview, 50 agreed to participate. In three cases, all arising in the relation to the same cluster, people declined stating that they had no interest in the project. As respondents indicated varying levels of support for their cluster, inactivity was not restricted to those who declined to participant and so is not thought to be influencing the results. When account is taken of firms that had closed, relocated or changed ownership since the membership list providing the sample population had been compiled, at least 50 percent of participants were interviewed except in the case of the Southern Wood Council. In the latter case, 9 of 21 members were interviewed but this excluded only 4 business members of which 3 were new recruits to the group at the time of the survey (late 2005).

At the outset some features of the forest industry need to be noted as they bear upon the results obtained. Timber processing is a nature-based industry (Prudham, 2002). This has implications for the extent of risk and uncertainty facing producers. Activity has an extensive geography (timber is harvested and transported over large areas); there are frequent changes in the terrain where logging occurs and to the specifications of timber harvested as well as variability in the weather. As in other land-based industries, this has traditionally resulted in much fragmentation of industry ownership to pass on the risk of production variability. In New Zealand, a feature of the industry is the survival of arm's length relationships between saw millers and timber suppliers. Timber suppliers prefer flexibility to long term relationships so as to exploit market instabilities as well as because of the potential variability in timber obtained from a single supplier. This context encouraged participation in separate industry associations and, at times, antagonistic relations between associations representing timber processing and those representing forest owners.