Budgetary constraints and programmatic choices by Flemish subsidized theatres

Abstract

We analyse programmatic choices of Flemish theatres and examine how they are affected by the theatres’ budgetary situation. Following Lancaster's characteristics approach, we identify several output characteristics of individual Flemish theatres during the period 1980 to 2000. A simultaneous equation approach is used to capture the theatre managers’, subsidizing government's and consumers’ behaviour. We find that changes in the budgetary situation of a theatre are translated into changes of both the ‘amount’ and the nature of the theatre's output. The budgetary impact on artistic choices has intensified since the introduction of a 4-yearly instead of yearly allocation of subsidies. The decrease in financial risk for the individual theatres leads to an increase in artistic risk-taking.

Notes

¹ His approach is in line with Niskanen's description of the budget maximizing bureaucracy (Niskanen, 1971). Note that nonprofit theatres are a particular example of a bureaucracy according to Niskanen's definition of bureaus as ‘nonprofit organisations which are financed, at least in part, by a periodic appropriation or grant’ (Niskanen, 1971).

² We consider the seasons 1980 to 1981 until 2000 to 2001. In the remainder of this study, theatre seasons will be referred to by the first year mentioned (for example, the season 1993 to 1994 will be referred to by the year 1993).

³ The selection of the characteristics was driven by the literature and the availability of data. Data on programmatic choices come from the ‘Vlaamse Theaterjaarboeken’ over the period 1980 to 1998, the database of the ‘Vlaams Theater Instituut’ and the Flemish Ministry of Culture.

⁴ An adaptation is an existing story or play that serves as the source of inspiration for a Flemish author who transforms it in an innovative way

⁵ Local governments–municipalities and provinces–also subsidize theatres. However, due to the large number of missing data, we did not take these subsidies into account.

⁶ An audience-maximizing manager is indifferent between programming n different plays with each one performance, or performing 1 production n times. Still, as programming a new production has sizeable fixed costs he will perform as few plays as possible (the constraint being that the play has to be popular enough to fill the theatre).

⁷ The number of subsidized theatres in any given year varies between 24 and 35.

⁸ The relevant market's income is an average weighted by population.

⁹ In practice, movie theatres are only located in (relatively) large municipalities and not in their neighbouring municipalities. Hence, using or not using the spatial weighting matrix makes no difference. Formally, however, the ‘relevant market’ spreads over to neighbouring municipalities even if in these municipalities no substitute (cinema) is present.

¹⁰ A specification with the lagged value of the budget was estimated as well. The results were similar to those presented in the main text.

¹¹ Following the current practice in empirical work on political budget cycles, we experimented with alternatives for DS. We defined a variable TUS _t which gives the ‘time until the following subsidy round’. So TUS _t  = 0 in all years before 1993 as well as in 1996 and 2000 (in the latter years decisions are taken on the subsidies for the following 4 years); TUS _t  = 1 in 1995 and 1999, 2 in 1994 and 1998 and 3 in 1993 and 1997. We could not find evidence of a ‘budget cycle’.

¹² A Hausman specification test examines the correct system specification, and determines whether 2SLS or 3SLS is the appropriate estimation technique. The results show that a simultaneous equations model is a better estimation technique (2SLS or 3SLS) than a single equation system (OLS), and that the full-information method (3SLS) provides better estimates than a limited-information method (2SLS).

¹³ This result seems to be in line with the finding by DiMaggio and Stenberg (1985) for US resident theatres. They find that resident theatres with a larger market have a less ‘conformist’ repertoire. Of course, diversity as measured by the number of different productions is not the same as ‘nonconformity’ as measured by DiMaggio and Stenberg's Conformity Index.

¹⁴ Alternatively, cast size expansions can reflect the situation where individual actors do no longer play multiple roles within one and the same play. While such a situation is not inexistent in Flemish theatres, it can by no means explain the empirical effect found in Table 2.

¹⁵ We have no straightforward explanation for this unless, maybe, that the main incentive to choose for new productions is to convince the subsidizing government to grant (more) subsidies. Indeed, as can be seen from column 9 theatres that are very innovative in these terms systematically receive higher subsidies. This incentive (to convince the subsidizing body) is likely to be smaller under the 4-year subsidy regime.