From Census Bureau surveys in 1972 and 1974, political scientists Wolfinger (Berkeley) and Rosenstone (Yale) have come to an array of significant, sometimes startling conclusions about U.S. voting patterns and voters' behavior. The single most important variable, they find, is education--with income and occupation distinctly subsidiary. Among occupation groups, farmers are the likeliest to vote--because of ""the variety of their relations with government,"" the authors suggest, and ""the wide fluctuations in their economic fortunes,"" which they attribute to government actions. Women overall are only slightly less likely to vote than men; but among elderly women, the decline in turnout is much Steeper than among elderly men (because women lose physical vigor sooner? because today's women were born into a male-dominated world? and these very women are widows?). Still, ""aging by itself produces not a decline but an increase in turnout."" Among the young, ""students vote at a much higher rate than non-students."" However, that increase in turnout among the aging is greatest for those with the least formal schooling--which suggests not only that the effect of education diminishes with age, but that experience is an effective teacher (both in ""coping with bureaucratic hurdles and thinking about political material""). Proceeding to deterrents, Wolfinger and Rosenstone note that ""registration is usually more difficult than voting""; and they project a 9. 1 percent increase in voter turnout were every state to adopt four liberalizing provisions (the most permissive in force anywhere in 1972), beginning with eliminating the closing date. If this were done, however, and the turnout rose--""the expanded voting population . . . would be remarkably similar to the actual voters."" In other words, and contrary to every assumption, the partisan political effect would be almost nil: Democratic candidates and liberal causes would not benefit. The reader may now anticipate the authors' capstone finding-that voters are representative of the populace as a whole (""The most likely to be under represented are people who lack opinions""). Some reassurance, then, for those disturbed by the low U.S. turnout, and a host of other implications that are sure to be discussed and disputed.