sector.100This fact might be an indication that the simple advice to spe- cialize in financial services and tourism is by far not enough to guaran- tee a high living standard. Note further that Liechtenstein, for example, which is often associated with its specialization in financial services, to our surprise has a smaller service sector than its neighbor countries Austria and Switzerland (see Section 5.2). It has to be admitted, how ever, that in Liechtenstein, the financial services branch constitutes a con si - derable proportion of the comparatively small service sector. We will come back to these important questions later on. Not unexpectedly, we observe high import ratios and export de- pendency ratios for virtually all VSC in our sample which comply with international economic theory and evidence. Furthermore, there seems to be a relationship between openness, as measured by imports as a per- centage of GDP, and per capita GDP, which would, again, be no sur prise (see, also, Section 5.2). Openness as a substitute to tiny national markets has already been mentioned as a strategy to overcome constrained growth paths. The export dependency ratio in Table A.14 rests crucially on the definition of neighbor countries, since that question is not a trivi- al one for island economies. Unfortunately, data are not available for all the VSC, but continental European VSC display an extremely high ex- port dependency. Without having accurate data on Liechtenstein, Mo - naco and San Marino, we are convinced that we are on the safe side in expecting their export dependency ratio as calculated in Table A.14 to be clearly above the figure of Luxembourg (61%). Such a high dependency reflects the political and economic stability in Western Europe and is – within a stable political and economic environment – not detrimental to VSC economies. On the contrary, it is supposed to be a prerequisite for a flourishing VSC economy. Unemployment, growth and inflation rates displayed in Table A.15 should be treated with caution, since data sources may not be consistent, base years may differ and figures refer to a single year without taking busi ness cycles, natural hazards etc. into account. Bearing these caveats in mind, it is however interesting to note that besides the enormous range of unemployment rates in VSC, growth rates (with the exception 131 
Economic characteristics of very small countries 100A correlation between per capita GDP and the proportion of the tertiary sector con- firms the notion of different consequences of specializing in services. The Pearson correlation coefficient is low (0.161) and insignificant.