Principal Component Analysis

#1
Are the sample variances of principal components always positively biased compared to population variances? I know that the first principal component has the largest variances and the last has the smallest, but how can i compare sample variances of them with population variances?

Thanks!