Donald Trump comes to the presidency with the lowest poll numbers of any incoming president since — well, since random sample public opinion polling started in October 1935. Pollsters haven’t been asking about Trump’s job rating yet, since as this is written he hasn’t as yet taken the job, but his favorables/unfavorables have generally been negative: 41 versus 50 percent in the Jan. 19 realclearpolitics.com average.
But his favorable numbers in particular have been widely different, 32 to 50 percent favorable, partly probably because of different question wording, partly perhaps because of particular news stories breaking on particular days.
We’ll start getting Trump presidential job ratings in a week or two. Here’s my rule of thumb for judging them: Adjust positive ratings 2 points upward and negative ratings 2 points downward. The reason, in a word: California.
As I pointed out in my Washington Examiner column in early December, California cast the second-highest Democratic percentage of any state in 2016; the numbers have slightly changed since then, because California for some reason takes five or six weeks to count its votes, but the final percentages show California voting almost exactly 13 1/2 percent more Democratic and 14 1/2 percent less Republican than the national average.
The national vote was 48 to 46 percent for Hillary Clinton, with California 61 to 31 percent for Clinton (with just a few more votes, each of those percentages would have rounded upward, making the score 62 to 32 percent for Clinton). In comparison, the other 49 states and the District of Columbia together voted 48 to 46 percent for Donald Trump.
As I pointed out in the column, this is the first time that the nation’s largest state has voted so lopsidedly at one end of the partisan spectrum. New York, the largest state from the 1810s to 1963, seldom voted more than 5 points off the national average, and neither did California from 1963 up through and including 2004. Since then the divergence has widened to 9 points in 2008, 10 points in 2012 and 13 points in 2016.
California’s increasing Democratic margins have helped Democrats win the popular vote. But that shift has made almost no difference in the electoral vote. Bill Clinton carried California by 13 points twice in the 1990s and got 54 electoral votes each time. Hillary Clinton carried it by 30 points in 2016 and got 55 electoral votes, the additional 1 coming from the reapportionment of House seats following the 2000 Census.
Everyone knows Trump or another Republican nominee will lose those 55 electoral votes in 2020 and, if California gets an additional electoral vote from reapportionment following the 2020 Census, that will not increase its number of electoral votes until 2024. So California’s popular vote does not change anything in the Electoral College.
Yet because California casts 10 percent of the nation’s votes and has 12 percent of the nation’s population, it will account for 10 to 12 percent of the respondents in national polls. Lopsided partisan preferences in small states like Vermont and Wyoming don’t tilt the national polls significantly one way or another, and tend to cancel each other out. The lopsided partisan preference of California voters tends to tilt national polls toward the Democrats by a perceptible amont — roughly 2 percent — in a way that is, as we learned (painfully, for Democrats) this year, irrelevant to which candidate gets a majority of electoral votes.
The traditional rule of thumb is that a president who gets 50 percent or more job approval is in good shape for re-election. Given the California tilt, my rule of thumb will be that if Trump is getting 48 percent or better he’s in good shape for re-election. In other words, don’t take the polls at face value.
