4.3.2 Theoretical expectations versus reality – defense Table A.10 in the Appendix provides evidence of the defense and ex ter nal security of VSC. Defense-related issues are considered to be a prime example for a public good. We, therefore, expect VSC to avoid sett ing up their own army whenever possible. The evidence for defense is not as clear as for the monetary system. seven of the selected VSC do not operate an army. Their external secu - rity is mostly guaranteed by a larger neighbor or regionally important country. The VSC without military forces are namely: Andorra (France and Spain guarantee security), Iceland (USA), Liechtenstein (Switzer - land), Micronesia (USA), Monaco (France), Nauru (Australia) and Palau (USA). There is no VSC with neither an army 
norany kind of security guarantee, although the latter may be more or less formal. Nauru, e.g., has only an informal agreement with Australia. Since most of the VSC mentioned above are of strategic interest for the defense policy of the security-guaranteeing countries, they can quite easily free ride on defense issues. Even if formal treaties exist, the «charge» for providing external security is supposed to be low or zero. Especially the marginal costs of providing security for VSC are approxi- mately zero. When you think of Liechtenstein for instance, it is obvious that the additional burden on the Swiss army for providing security is negligible, at least in times of peace. There is, of course, a disadvantage for islands with regard to de - fense provision, because they generally do not lie as close to a larger country’s immediate strategic interest sphere as landlocked countries. Therefore, they are expected to have a higher probability of operating their own army. With the exception of Iceland, Micronesia, Nauru and Palau, indeed, all island VSC corroborate this conjecture. The only two countries where an army was to be expected, are Belize and Brunei. For both, the preference homogeneity with adjacent countries is – as also ex- pounded above – relatively low. In general, theoretical expectations are not met in Table A.10. Twelve out of 21 VSC were wrongly presumed to not have their own army. The result seems more in line with Hypothesis 1 when we account for several additional features. With the exception of Brunei, which has very high relative expenditure for its army, all other VSC for which data were available, do not spend more than an average larger country in per- 94 
Very small countries: organizational choice and international outsourcing