To what extent does religious affiliation determine how Americans vote on election day? Menendez' answers to this vast question are not astounding. Except for rabid sectarian enclaves, religion generally plays a marginal role. There are exceptions: Catholics vote for aid to parochial schools, Jews and Protestants against. Yes, aversion to popery helped defeat Al Smith in 1928, but scholars are not agreed whether Rum or Romanism hurt him most--or even, whether any Democrat could have beaten Hoover. JFK won in 1960 despite a net loss of 1.5 million votes due to his religion--according to a Univ. of Michigan study. One isn't shocked to learn that ""the religious left favors the political left,"" or that high-status Episcopalians are more likely to vote along liberal lines than Southern Baptists. For those who have the patience, Menendez has checked electoral results for '68 and '72 in select, religiously homogeneous counties and discovered, for example, that ""German Catholic Republicans switched en masse to Kennedy."" The religious complexion of Congress, Menendez finds, shows a consistent under-representation of Lutherans (no Lutheran has been a presidential candidate), a phenomenon he attributes to dour Lutheran theology and a willingness to leave politics to Caesar. Not a very satisfactory explanation. As to the resurgence of Evangelicism, now that Carter is in the White House--it's too soon to say what this portends. (An epilogue breaks down religion at the polls in the '76 election.) The most arresting scrap of information comes from Rush Springs, Okla. (pop. 1,000), which recently voted a ban on both public and private dancing.