Why do we use n-well in p-substrate for CMOS technology instead of using p-well in n-substrate?

What are the advantages of using a p type substrate then n type substrate in CMOS circuit implementation?