This is a stupid conversation but just so someone cites actual data and not just opinion slapfighting:

Ban

In 1982, President Reagan created a national commission on drunk driving which resulted in several important recommendations that would become foundations to the U.S. approach to stopping drunk driving. The commission issued a report in 1983 which called for raising the minimum drinking age to 21 and for tough enforcement of drunk driving laws. src

… there has been a 38 percent drop in drunk driving deaths since 1982. src

Education

Laws aimed at alcohol-impaired driving have been shown to change behavior in ways that reduce the problem. Alcohol education and public information programs, in contrast, rarely result in short-term behavior change. In part, this is because drinking, and combining drinking with driving, are lifestyle behaviors shaped and supported by many ongoing social forces, and they are not readily amenable to change through brief, one-time education/public information efforts. Moreover, those who contribute most to the problem have characteristics that make them least susceptible to behavior change through educational programs. However, education and public information programs have an important role to play in combating alcohol-impaired driving. They can provide support and impetus for passing laws; transmit knowledge about the provisions and penalties of laws in ways that increase their deterrent effect; and generate public support for law enforcement programs. src (emphasis mine)

In contrast, an education program that research has shown to be effective simply refers back to the ban itself in the first place, i.e., the You Drink, You Drive, You Lose program was successful, and was focused around informing people that DUI activity will be caught and punished. https://uknowledge.uky.edu/ktc_researchreports/244/

In summary, chill out. Both bans and education have contributed to the improvement we see today and your narrative that bans are conservative and somehow ineffective is so easily refuted by the data.