Back in the old days, when I was a political
campaign manager (early 1960s), I started a group in San Francisco
called Research in Politics. One of the experiments I did was to find
out if election day "get out the vote" worked.
I got the Republican and Democratic Party organizations to carry out a research project to find the answer. I picked two special elections for Congress [Phil Burton (Dem) in San Francisco and Jack Veneman (Rep) in Modesto] because a special election only has one issue and if "get out the vote" could ever work it would be in a special one ballot issue election.
In both districts I randomly chose 25 precincts and divided them into five groups of 5 precincts. We had one group of 5 precincts that got 100% Republican effort no Democratic workers, one group of 5 that got 100% Democratic effort with no Republican workers. There were two groups of five precincts that got 50% effort from the Republicans and another two groups that got 50% effort from the Democrats. One last group of 5 precincts that got absolutely no "get out the vote" effort.
Then we sat back and watched the results.
"Get
out the vote" efforts had absolutely no effect...none. The voter
turn-out in each group (in both campaigns, 10 groups with fifty
precincts) was the same for both parties. No effect of "get out the
vote."
I was even flown to Chicago to a National Republican training session to present the results along with a few other research projects I did.
Outcome: Both parties understood the meaning of my research, but both parties decided that "get out the vote" was such an important way to make use of campaign volunteers that the results didn't matter.