The arts in many countries, but particularly in developing ones, are coming under increasing financial pressure and finding it difficult to justify the increases in government funding needed to maintain and grow the cultural sector. The trend in cultural economics, as well as in other areas, appears to be towards including qualitative valuations, as well as the more traditional quantitative ones. This thesis argues that the value of cultural events should include long term historical qualitative analysis, financial or economic impact and a valuation of the positive externalities provided by cultural events and that any one of these should only be regarded as a partial analysis. Four methods of valuing the arts using the South African National Arts Festival (NAF) as an example are demonstrated. Firstly, a qualitative historical analysis of the role of the NAF in South Africa’s transformation process from Apartheid to the democratic New South Africa is examined, using theories of cultural capital as a theoretical basis. It is argued that the value of cultural events needs to take into account long-term influences especially in countries undergoing political and social transformation. The second valuation method applied is the traditional economic impact study. Four economic impact studies conducted on the NAF are discussed and methodologies compared. It is concluded that, despite the skepticism of many cultural economists, the method can provide a useful partial valuation and may also be used for effective lobbying for government support of the arts. Chapter four discusses willingness to pay studies conducted at the NAF in 2000 and 2003 (as well as a pilot study conducted at the Klein Karoo Nationale Kunstefees). It is found that lower income and education groups do benefit from the positive externalities provided by the Festival and that this is reflected in their willingness to pay to support it. It is also argued that such contingent valuation studies can provide a reasonably reliable valuation of Festival externalities, but that they may be partly capturing current or future expected financial gains as well. Finally, the relatively new choice experiment methodology (also called conjoint analysis) is demonstrated on visitors to the NAF. The great advantage of this method in valuing cultural events is that it provides part-worths of various Festival attributes for different demographic groups. This enables organizes to structure the programme in such a way as to attract previously excluded groups and to conduct a cost-benefit analysis for each part of the Festival.

Over the course of the last three decades, it has become popular practice to evaluate tourism events like cultural festivals in financial terms, through the use of economic impact studies. This can be attributed at least in part to the notable growth in the number of festivals being held globally and, as such, a higher level of competition between festivals for the limited funding which is available. Economic impact studies, and the resultant findings, have thus become powerful tools for the lobbying of sponsorship, and it has become increasingly important that the impact calculations be as accurate as possible, so as to effectively allocate both government and private resources to projects which will be of the greatest benefit to the host region. The allocation of funding is especially vital in an area like the Eastern Cape Province of South Africa, which is faced with many financial difficulties. The allocation of public funds to an event like the National Arts Festival, which is hosted in a relatively wealthy part of the province, might be weighed against initiatives which directly benefit the poorer parts of the region. Although it is acknowledged that the benefits which are felt by the host community of a cultural event go beyond that of the financial, it is often on this basis that festivals are most easily compared. The primary goal of the thesis was to analyse the various forms of methodological bias which can exist in the economic impact analyses (EIA) associated with cultural events. Theoretical considerations were discussed, specifically regarding economic impact as a method of measuring value. Various forms of bias (including data collection, the calculation of visitor numbers, multipliers, defining the area of interest, inclusion of visitor spending, and accounting for benefits only, not costs) are put into a real-life context, through the investigation of economic impact studies conducted on three selected South African festivals (the Volksblad, the Klein Karoo Nasionale Kunstfees, and the National Arts Festival), and one international festival (the Edinburgh Festival). An in-depth comparison of two separate studies conducted at the National Arts Festival (NAF) in 2004 (by Antrobus and Snowball) and 2005 (by Saayman et al.) was made, focussing on the manner in which the economic impact was calculated. Having considered the common forms of bias, and assessing several possible reasons for the difference of approximately twenty million Rand in the advertised economic impacts, it was concluded that, most likely, the miscalculation of visitor numbers was the cause. This was confirmed when the Antrobus and Saayman methods were applied to the 2006 NAF data, and noting that the economic impact figures arrived at were strikingly similar. As such, it is advisable that extreme caution be taken when calculating visitor numbers, as they can significantly influence the outcome of an economic impact study. It is recommended that each study should also have transparent checks in place, regarding the key calculation figures, to ensure that less scrupulous researchers are not as easily able to succumb to the pressure event sponsors might impose to produce inflated impact values.