Methodology

Information gathering comprised several stages, which are enlarged upon in the following sections. The analysis framework (Section 2.1), which forms the core of this study, was largely based on the work of Schmithüsen, Bisang and Zimmermann (2001). Several additions were made prior to and during the classification of examples when limitations became apparent. The framework is presented and described in Section 2.1. Participants were selected primarily from within FAO but also outside, on the basis of their involvement with policy and/or planning in areas related to development and natural resource management (Section 2.2). Examples of occurrences where 1. public policy positively or negatively affects sustainable forest management, or 2. forestry policy positively or negatively affects other areas were gathered through interviews and also by use of an e-mailed questionnaire (Section 2.3). Following an initial assessment, further research into each example was undertaken using the Internet and literature provided by participants (Section 2.4). This was necessary in order to provide additional information corresponding to the requirements of the analysis framework. A planned follow-up with participants to estimate the scale of effects and the accuracy of final details was not undertaken due to time constraints although the resulting details were sufficient for a preliminary analysis (Section 2.5).

Several of the issue encountered in developing a methodology to analyse cross-sectoral impacts in forestry related to definitions and their application. The question of whether to include examples of public policy affecting forests but not forestry arose through application of the title of this work: How can a cross-sectoral policy impact occur where there is no involvement of a formal forestry sector? However, because of a) the potential non-market benefits of forests to diverse groups of people and b) the difficulty of separating forests and forestry where forest activities are informal, examples were included.

A number of examples were given where forests or forestry were being affected not through pursuit of a particular policy but through normal activity. This posed the question of how activity that may run contrary to the aims of sustainable forest management should be treated. Without a thorough analysis of the responsibilities, goals and policies of all forestry related sectors it was considered that inclusion of such situations would pose too great a problem for the current study. The decision was therefore taken to exclude these examples from the analysis but to include them in the catalogue. By doing so the extant situation was taken as a baseline rather than the principles of sustainable forest management which becomes difficult to define with the temporal and spatial specificity of particular examples. The issue of what constitutes ‘forestry policy’ arose in light of the fact that forestry departments often enact policies with aims outside of the traditional forestry mandate (e.g. watershed protection). A decision was taken to include all policies carried out by forestry departments as forestry policy and thus to accept the extant situation as the baseline, in similarity with the preceding case. This has the effect of making cross-country comparisons more difficult, given that forestry constitutes different things in different countries. In connection with this situation several examples were excluded from this study as they were effectively forestry policies that affected forestry itself. This posed problems as it was recognised that the aims of the forestry policies were often far from what would usually be considered forestry, being more closely allied to, for example, short term income accumulation by individuals.

Finally, integrated policies that contained forestry were also difficult to include as they were often too recent to determine whether forestry was efficiently managed within the integrated policy framework or otherwise. Moreover, integrated policies are already seen as desirable from the point of view of accounting for cross-sectoral policy impacts and were not therefore a primary focus of this work.

The forest/socio-economic categories given in the table were taken directly from Schmithüsen, et al (2001) having been proposed by de Montalembert (1994) for their utility in examining cross-sectoral linkages. High/low population density was ascribed according to a boundary of 65 persons/km² in 2000, the median for all countries taken from the UN 1998 revision of World Population. The classification follows that proposed by the IPF/IFF/UNFF secretary, Jag Maini. The ‘low forest cover country’ classification accorded with the list produced by the International Forum on Forests, but was not used due to the small number of examples from within these countries. The high/low income boundary used was US$ 2995 GNP in 1997 which accords with the boundary between lower and upper-middle income countries according to the World Bank’s classification. The high/low forest cover boundary was 20% according to the figures given in the FAO Forest Resource Assessment (2000).

The breakdown of public policy was again taken from Schmithüsen, et al (2001) being originally derived from von Prittwitz et al (1994). Several additions were made; in particular the agriculture category was split into two categories, cash crops, grown primarily for foreign exchange earnings and subsistence/commercial crops grown primarily for home consumption.

Areas affected by external policy were categorised to broadly capture principle themes as follows:

Forest cover (deforestation/afforestation),

Forest alteration (including biological diversity, productive capacity, health and vitality of forest ecosystems)

Management capacity/application of policies

Processors of forest products

Forest products supply during term of policy activity,

Soil and water resources

Socio-economic effects for local people and forest users

The groups within each of these main categories were chosen on the combined basis of those used by Peck and Descargues (1997) for analysis of the European forestry sector (forest resources, wood supply, wood-processing industries, international trade in forest products and markets and demands for forest products) and the criteria developed for the ‘Montreal Process’ on criteria and indicators for the conservation and management of temperate and boreal forests. The latter, although aimed at developed countries, bear close resemblance to the criteria defined by the International Timber Trade Organisation for tropical countries.

Assessment of the scale of impacts was intended to provide a basis upon which estimates could be further quantified by combination with published figures, where available, on the amount/size the value concerned (e.g. forest cover within the area affected by a policy). Assigning scores required an assessment of the responsibilities and goals of the sectors concerned and therefore in some ways imposes an ideology. Scoring, however, was intended to show impact of the policy on given areas rather than promoting a view of what ‘should’ happen. In the analyses presented in Section 3 the scores were not finally used and remain demonstrative. However, the positive/negative nature of effects were used and were intended, as far as possible, to indicate the direction of effects of a policy on an existing quantity rather than whether good or bad.

Although some indication of the term length of policy impacts and the period of activity of a policy should ideally be included in the framework time and availability of information prevented this extension. In lieu of this, all impacts were assessed with awareness of the impacts in the short term and the focus of the work was on current or recent policies.

The policy decision level classes and the categories of actors involved in decisions making were based on those of Schmithüsen, et al (2001) and finally the categories policy instrument were from Baumol and Oates (1993).

Table 2.1 Analysis framework giving the example of deforestation resulting from roading and land use policies in Brazil and Central America Forestry being affected by an external policy

For examples of external areas being affected by forestry policy the above framework was modified by removing the policy areas involved and exchanging the classes of area impacted to include the four listed below.

Other forest users

Other special economic sectors (excluding soil and water resources)

Soil and water resources

Natural environment

These were chosen after review of the examples given and were thus limited by the relatively small number of examples given and need to develop a meaningful summary. They were not intended to delineate the full range of possible effects.

Strictly, the sampling unit in this survey was an intersectoral impact. To include a representative sample of the global ‘population’ of impacts would, however, be an impossible task given the fact that impacts are numerous, often ubiquitous and poorly defined. Therefore a survey of experts was used and although the study was intended to be exploratory in nature it is helpful to look at some of identified sources of bias such as:

Thus, a considered approach to interpretation of the results should be taken and conclusions need be cautious.

Interviews

Interviews were arranged in person following the dispatch of a message detailing the study. All interviews were carried out by the author between July 16^th and August 28^th. A simple, unstructured format was used, based on following question:

We are looking for present or recent examples of instances where policy, outside the forestry sector, influences forestry in either a positive or negative way, or of influences where forestry policy affects other areas in either a positive or negative way.

Where this question and any necessary clarification failed to elicit a response the following examples were given to show the broad range of possibilities and to ‘jog memory’:

Watershed management policies, which encourage afforestation to improve water quality/supply or where forest management techniques damage water quality.

Policies to subsidise roading and livestock such as have affected forestry in South America.

WTO agreements to lower of tariffs and trade barriers which may affect forestry.

Areas which had been noticeably absent during the course of the interview were focussed on towards the end of the interviewing schedule in an effort to gain an extensive coverage.

Questionnaire

A copy of the questionnaire to contacts outside of FAO headquarters is shown in Appendix 2. A simple format was chosen for two reasons, 1. The likelihood of ‘questionnaire fatigue’ amongst potential participants, 2. initial interviews had shown that highly detailed policy information was often not available from participants. The questionnaires were sent on 1^st Aug with a reminder and another copy of the questionnaire three weeks later.

Insufficient time was available to research examples of impacts where publications were referred to instead of examples themselves. Amongst others these included:

In most cases the detail of information required to complete the analysis framework (Section 2.1) was greater than that immediately available from the participants. Therefore, literature provided by participants and the Internet were searched for additional information. Problems were encountered with the lack of information on certain areas, such as the actors involved in policy making decisions and the policy instruments used. This results from the focus of published information on the effects of policies rather their origins which are naturally of more of academic interest.