Proceedings from the ECFIN Workshop "The budgetary implications of structural reforms" - Brussels, 2 December 2005

2 The arguments and literature survey

There is a small but growing literature aiming at identifying the drivers of structural reforms. With
a focus on financial deregulation, Abiad and Mody (2005)43 have developed a model describing
deregulation processes as a lagging adjustment characterized by status quo preferences towards
some target level. Helbling et al. (2004), Annett et al. (2004) and Heinemann (2005) have applied
this deregulation model to a larger spectre of reform issues including labour markets, product
markets, tax and trade reforms. Approaches abstracting from deregulation dynamics but
concentrating on explaining the likelihood of reforms under given economic and political
circumstances with the help of probit estimations are Pitlik (2003), Heinemann (2004), Duval and
Elmeskov (2005), Belke et al. (2005) and Deroose and Turrini (2005). Different reform indicators
and different potential drivers of reforms are being tested in this literature including proxies for
the state of the economy, the degree of globalization, political-economic factors or the exchange
rate system. A number of these studies also discuss the role of the budgetary situation and include
fiscal variables among the factors relevant for a country’s ability to achieve structural change.

2.1 The arguments

For summarizing and supplementing the main arguments of the literature and as a conceptual
framework for the empirical testing it is helpful to work with the following stylized time scheme
of a reform cycle. 44

Table 1: Stylized scheme of a reform cycle

It can be assumed that the link between the reform event in period R and the resulting long-term
fiscal situation (period long-term after reform: LA) is favourable. Basically, this positive link is of

⁴³ The earlier IMF Working Paper version dates back to 2003.

⁴⁴ It is stylized since it is implicitly assuming a reform process to be a one shot event and not a possibly continuous process as it will be modelled in
section 4.

133

<<   <   136   137   138   139   140   141   142   >   >>