Regional development goals and distributive politics in the allocation of Turkey's central investments: socioeconomic criteria, parties and legislators' personal networks

A growing body of theoretical contributions and empirical studies has explored how the spatial distribution of public resources and governmental transfers is likely to be driven by efficiency and equity, as well as by political considerations. This article applies this line of research to regional development policy making and explores the territorial distribution of public resources aimed at reducing interregional imbalances - namely fixed capital investments - in Turkey, a fast-growing, Mediterranean country currently candidate to EU accession. As early as 1963 Turkey established an ad-hoc institution and a specific policy agenda aimed at curbing the high regional disparities which, however, still dramatically persist. By adopting both Generalized Method of Moments and Fixed-Effects panel data estimators, the article first measures the extent to which socioeconomic criteria and developmental needs - i.e. the official policy principles - were central in the allocation of resources to Turkey's 81 provinces over 2002-2011. If the effects of special interest politics and interest groups on policy outcomes seem to be a fact overall ubiquitous in many political systems, further efforts are needed to reconcile competing models of distributive politics. Recent research suggests that who are the beneficiaries of pork-barrelling will depend on the type of electoral system and the strength of formal political institutions. Distributive predictions for proportional representation systems with weak parties and widely spread personalised consensus building practices, such as Turkey as well as other comparable emerging countries, remain however unclear. We thus test whether political pork-barrels most benefit formalised political groups such as parties or, rather, personalised networks of representation. While following the literature, we also include additional explanatory factors, namely same-hometown loyalties. The results of the analysis uncover that political factors matter, but less than socioeconomic rationales. In fact, even after controlling for political manipulations, socioeconomic disadvantage measures remain more relevant predictors of investments. Compared to the results of research carried out on Turkey's government grants and private investment incentives, public investments seem less prone to political manipulation. Contrary to what may be expected, however, results also suggest that it is mainly parties rather than individual legislators who benefit from redistributive games. Overall, the paper contributes to the assessment of regional development policies effectiveness in Southern current and candidate EU-Member countries, while also shedding more light on how much and who are the beneficiaries of spatial distributive games in regional development initiatives.