Hi,
you could take the country-wide proportion as a given and use one sample p-tests or just use two sample p-tests where one sample is always the one from the whole country. In both cases you would need to do some correction for multiple testing (like Bonferroni).
regards