Capital account liberalization and unemployment in industrial countries

Abstract

Using data on 21 industrial countries over 1973 to 2005, this article finds that the liberalization of capital accounts implemented during this period has probably reduced unemployment. The magnitude of the estimated effect is substantial. We control for both endogeneity of capital account regulation and all major determinants of unemployment. The results are robust to variations in specification.

Keywords:

• capital account liberalization
• capital controls
• labour market
• unemployment

JEL Classification:

• E24
• F36
• F40
• J64

Notes

¹ For recent surveys, see Henry (2007) and Kose et al. (2009).

² Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Japan, the Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, the UK and the USA.

³ The variable ranges from 0 (least restrictive) to 1 (most restrictive). It has been constructed by the author of this article using data from Ostry et al. (2009).

⁴ Trade openness is defined as the ratio of exports and imports of goods and services to GDP; source: World Bank (2011).

⁵ The unemployment rate is defined as unemployed as a percentage of the civilian labour force (harmonized rates); source: OECD (2011).

⁶ Sources: trade union density, collective bargaining coverage, wage bargaining coordination and centralization – Visser (2011); tax wedge – Nickell (2006); unemployment benefits replacement rate, employment protection legislation, product market regulation, output gap – OECD (2011); and real interest rate, inflation rate – World Bank (2011). For definitions, see these sources. Wage bargaining coordination and centralization, employment protection legislation and product market regulation are scaled to range from 0 to 1. The other variables mentioned in this footnote are in decimal fraction.

⁷ GDP per capita is in thousands of US dollars, constant prices and PPP; source: OECD (2011).

⁸ Source: World Bank (2011). The shock variables have been constructed by the author of this article as the difference between the actual and the smoothed indices as a decimal fraction of the smoothed indices, the latter calculated using the Hodrick–Prescott filter (λ = 6.25). The FDI variables are in decimal fraction of GDP.