Determinants of the link between financial and economic development: Evidence from a functional coefficient model Helmut Herwartz, Yabibal M. Walle ⁎ Department of Economics, Georg-August-University Göttingen, Germany abstract article info Article history: Accepted 20 November 2013 JEL classification: C14 C33 O16 G28 Keywords: Finance–development relationship Financial development Economic development Functional coefficient model Noting that “one size does not fit all” in the case of the finance–development (FD) relationship, a growing body of literature has recently focused on uncovering economic conditions under which financial development could be beneficial (detrimental) to economic development. We look into these conditions by means of a flexible semiparametric approach that allows the long-run FD link to depend on measurable economic factors. Using an- nual data for 73 economies spanning the period 1975–2011, we find that the impact of finance on economic de- velopment is generally stronger in high-income than low-income economies. However, allowing for intra-group variations reveals the importance of other factor variables in explaining the FD link. For instance, increasing fi- nancial development strengthens the FD link while increasing government size weakens it. Moreover, the FD link could even be negative if low- and lower-middle-income economies have very large governments or are ex- tremely open to international trade. © 2013 Elsevier B.V. All rights reserved. 1. Introduction The importance of services and instruments of the financial system to the real economic sector has been recognized in the literature at least since Schumpeter (1911). However, there are economists who argue that finance does not matter to economic development. Accord- ing to this view, either the financial system passively responds to the de- mand arising from the real sector and not vice versa (Robinson, 1952), or there is not at all a meaningful relationship between financial and economic development (Lucas, 1988). The intensive research on the link between financial and economic development in the last two decades has documented mixed results. 1 Empirical studies on the relationship between financial and eco- nomic development mostly follow either of the following two research directions. The first group of studies attempt to test whether financial development matters for economic development—independent of an eventual reverse causal impact (e.g. King and Levine, 1993; Levine et al., 2000). While such studies try to immunize estimations of the impact of finance on development from potential biases induced by reverse causation from economic to financial development, they gener- ally neither test the presence nor estimate the strength of this reverse causation. Often using cointegration and Granger causality tests, the second group of studies, however, explicitly examine the direction of causality between financial and economic development (e.g. Ang and McKibbin, 2007; Demetriades and Hussein, 1996). As an extension of the first research avenue, a growing body of literature has recently attempted to investigate underlying measurable economic conditions (henceforth factors) which might determine the impact of finance on economic development (henceforth the finance–development (FD) nexus/link/relationship). 2 These studies raise a question of substantive policy relevance: under what conditions is financial development Economic Modelling 37 (2014) 417–427 ⁎ Corresponding author at: Georg-August-University, Platz der Göttinger Sieben 5, D- 37073 Göttingen, Germany. Tel.: +49 551 39 7671; fax: +49 551 39 7279. E-mail address: ywalle@uni-goettingen.de (Y.M. Walle). 1 For instance, a significantly positive impact of financial development on economic de- velopment is documented in Christopoulos and Tsionas (2004), King and Levine (1993) and Levine et al. (2000). However, there are also studies which report that it is economic development which leads to financial development (Ang and McKibbin, 2007). In addi- tion, a few studies have diagnosed a negligible impact of finance on economic develop- ment (Andersen and Tarp, 2003). See Levine (2005) and Ang (2008a) for extensive surveys of the theoretical and empirical literature on the relationship between financial and economic development. 2 The two groups of studies use the phrase “FD nexus/link/relationship” with slightly different meanings. In the first group, as in this study, it means “the impact of finance on economic development”. In the second group, however, it more broadly refers to “the (causal) relationship between financial and economic development”. Similarly, the term “growth” is often used in the literature together with, or instead of, “development” even when it does not refer to the “rate of change” of income. In particular, studies focusing on the long-run FD relationship (e.g. Christopoulos and Tsionas, 2004; Demetriades and Hussein, 1996) use the level of real GDP per capita to measure economic “growth” or “de- velopment”. It should be noted, however, that empirically distinguishing between “growth effects” and “level effects” of growth determinants is both complicated and less important as we are eventually concerned with improvements in welfare levels (Temple, 2000). Nevertheless, since the dependent variable in this study is GDP per capita, we pre- fer to use the term “development” instead of “growth”. We thank an anonymous referee for encouraging us to deviate from the literature in this regard. 0264-9993/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.econmod.2013.11.029 Contents lists available at ScienceDirect Economic Modelling journal homepage: www.elsevier.com/locate/ecmod