"History doesn't repeat itself, but it does rhyme."
Mark Twain
This is Labor Day weekend, the traditional kickoff for a national campaign — but, like so many other facts of political life in the United States, that has been rendered essentially invalid in 2016. This campaign never took a break, not even between the conventions and the Labor Day holiday, which is how it has been in the past.
Before a serious evaluation of the campaign could be made, it was necessary to allow some time to pass after the conventions were over. It has now been more than a month since the conclusions of the conventions, and it is now appropriate to look at the polls and see what they can tell us.
But the polls are apt to be volatile until after the debates so my advice is to treat the polls from now until about mid–October as forms of entertainment.
We heard a lot of talk about
making history during the Democrats' convention, just as we did eight years ago when Barack Obama became the first African–American to be nominated for president by a major political party.
There are always emotional appeals in political campaigns, but there are many other factors that drive campaigns, too. A presidential campaign is long and often buffeted by sudden, unforeseen events. This year's campaign, with historically high negative ratings for both major party nominees as well as a general election campaign season that is longer than usual simply because both conventions were held in mid–summer, seems particularly prone to week–to–week, if not day–to–day, fluctuations.
I tend to disregard the polls immediately after the conventions because candidates nearly always get a post–convention
bounce. They should. Since the advent of television, parties have been refining their conventions into the choreographed four–night propaganda fests they have become.
If a candidate doesn't get a bounce, even a small one, the convention planners did not do their jobs. It usually takes something like riots in the streets (see
Chicago in 1968) to prevent a candidate from receiving a bounce.
In the aftermath of the Democrats' convention, I heard many Democrats boasting of Hillary Clinton's nine–point bounce in the polls. Historically that is about average. It is what George W. Bush received after being re–nominated in 2004, but that margin didn't hold up. He went on to win but by a narrow margin over John Kerry.
To me it suggests there is a sizable portion of the electorate that remains suspicious of Clinton — and wasn't persuaded by the Democrats' Hillary lovefest/Trump bashfest.
Such bounces tend to be short–lived as Americans' ever–shrinking attention spans shift to other things. Al Gore got a double–digit bounce in 2000. Ditto Michael Dukakis in 1988. Look up those two in your history books. In no history book will you find either man having been sworn in as president.
I tend to take polls more seriously the closer we get to the election itself — or, in an era when many voters can cast their ballots up to a month before the actual Election Day, the closer we get to the start of early voting. By October, the debates will have begun, and many voters will have started casting their ballots (as I understand it, a few voters have already cast their ballots). That's when the polls will begin to reveal what we can expect in November.
The polls of October will reflect events that haven't happened yet and the candidates' responses to them. They will have more relevance to the election. They will dictate where campaign resources are allocated.
The old rule of thumb was that people didn't start paying attention to the campaign until after the World Series. In some places, I suppose, that timetable has been moved up a few weeks, maybe to the start of the NFL's season in September.
The polls in August are snapshots of the start of a horse race and should not be considered predictive in any way, but they
can be useful as analytical tools, and they can demonstrate convincing trends to expect on Election Night.
The opportunity to make history isn't always enough to win an election. Yes, Obama went on to win, but neither Geraldine Ferraro, the first woman on a major party ticket, nor Sarah Palin, the
second woman on a major party ticket, won; Al Smith was the first Catholic to top a major party ticket, but it was John F. Kennedy more than 30 years later who became the first Catholic to be elected president.
We've had Catholics on other national tickets since. Some have won, some have lost. It isn't a subject that is mentioned anymore, and that is probably the chief value that being first provides. Being a black or Catholic — and now, a female — nominee for president no longer raises any eyebrows.
Mitt Romney was the first Mormon to be a major party's presidential nominee, and Joe Lieberman was the first Jew to be on a major party ticket, but neither won. Being first doesn't ensure success.
But it
is important, in a country that prides itself on being the land of opportunity, that things like race or gender or religious faith don't get in the way of participation in the process. The voters will then decide if a candidate is qualified for the office he/she seeks.
Make no mistake about it, Clinton's nomination is significant symbolically; win or lose, she has earned a spot in the history books. The voters will decide the rest, and they will make that decision after taking into account those things that
they think are important.
I'm sure the symbolic nature of Clinton's nomination will influence
some voters — although I suspect most of those voters would have voted for Hillary, anyway.
Politicians don't get to tell voters what is important and what is not. They can say what they think, but they cannot insist on what can be considered and what must be disregarded. Judges can do that with juries; politicians are not permitted to do that.
A presidential election is a complex thing, anyway — starting with the fact that it is really 51 elections with 51 sets of issues that are important to different sets of voters with the electoral votes from each state riding on the outcomes. When voters go to the polls, they may
think they are voting for Candidate A or Candidate B — but they are really voting for a slate of electors who will represent the candidate's party in the Electoral College.
Most of the time, those electors support the nominee of their party — but not always. Such electors are called
faithless electors — but that is a subject for another time.
(In hindsight, Obama's election may seem inevitable to those who don't remember that, until the economy imploded in mid–September of 2008, Obama was trailing John McCain in many polls — and none other than his running mate,
Joe Biden, once mused in public that Obama should have picked Hillary Clinton, his runnerup in the primaries, to be his running mate in the interest of party unity.
(Biden's selection had been rationalized as Obama's attempt to make up for inexperience in foreign policy in the wake of escalating tensions between former Soviet republics Russia and Georgia in the summer of 2008. Ironically, after he was elected, Obama chose Clinton to be his first secretary of state.)
The one thing I have taken from poll after poll is that there is considerable fluidity in this year's electorate; consequently, as a lifelong student of history, I find it more relevant to observe what American voters have done when a term–limited president has had to leave office and those voters have had to select a new leader. I have often heard it said that such an election — in which the incumbent is prohibited by law from running again — is a referendum on that incumbent's performance — and, after eight years, voters generally are ready for a change (I guess you could call it the
eight–year itch).
If the rise of Donald Trump on the Republican side and Bernie Sanders on the Democratic side proved nothing else, they demonstrated that there is a considerable desire in this country for a different direction, something the polls have supported consistently. There is just disagreement about which direction to take. To — kind of — quote Howard Beale in
"Network," the voters are mad as hell and they're not going to take it anymore.
And such people are liable to do anything.
Conventional wisdom was turned on its ear this year as a billionaire novice politician and a 74–year–old socialist took their parties' nominating processes into uncharted waters. If you think there are no surprises left, let me remind you that there are more than two months left until Election Day.
Under these circumstances, it is hard to get a handle on what to expect. A lot of people these days say Clinton will win in a landslide — but, as far as I can see, their predictions are based almost entirely on the polls that have been coming out since the Democrats wrapped up their convention four or five weeks ago. Already we are seeing signs of the race tightening in some states. While it may yet wind up being a blowout, I am still inclined to believe it will be close. External factors — those peace and prosperity metrics — simply are not what have been historically required for the incumbent party to win without the incumbent topping the ticket.
Presidential term limits have existed since the ratification of the 22nd Amendment 65 years ago. Since that time, four presidents (Dwight Eisenhower, Ronald Reagan, Bill Clinton and George W. Bush) have served two four–year terms. Another (Richard Nixon) was elected to two terms but resigned less than halfway through his second term. Thus, Obama is the sixth president to be elected to two terms since the ratification of the 22nd Amendment.
Three of Obama's postwar predecessors who were elected to and served two terms saw their parties lose the White House when they were forced by law to step down as Obama is today. The one exception to that was when Vice President George H.W. Bush was elected to succeed Ronald Reagan in 1988.
It is tempting to chalk that up to Reagan's popularity — and I am sure that played a role in Bush's victory that year.
But I don't think approval ratings tell the whole story.
Let's compare Reagan, as the only postwar two–term president who was succeeded by someone from his own party, to the others. A useful comparison point is the December of the president's seventh year in office — less than a year before his successor was chosen. Reagan's approval rating in December 1987 was 49%, which was better than Bush 43 (30%) in December 2007 but not as good as Clinton (55%) in December 1999. Reagan's approval rating also bested Obama's in December 2015 (46%).
If presidential approval was the only determining factor, Clinton's vice president, Al Gore, should have beaten George W. Bush in 2000 — and, indeed, he
did in the popular vote but
not in the electoral vote that has decided the outcomes of presidential elections since the 18th century. The prolonged legal battle over the state of Florida was won by Bush, giving him enough electoral votes to win. If Gore had carried Tennessee, the state both he and his father represented in the U.S. Senate, or West Virginia, which had not voted for a nonincumbent Republican since 1928, Florida would not have mattered.
Another example of the complexity of presidential elections.
It's important to mention at this point that political science really isn't a
science at all — I figured that out when I was a college freshman studying political science. There is no formula in which you can fill in the blanks on each candidate's experience, knowledge and shortcomings and determine which candidate will win. The voters get to decide which factors matter the most to them, and that can — and does — change from one election to the next.
What's more, the United States is a relatively young country compared to most, and there really isn't that much of a voting history to study. There have been fewer than 50 presidential elections since they were opened to the people. (Until 1828, the only participants in presidential elections were the voters in the Electoral College — a handful of men at best from most states whose choices, it should be noted, were generally pretty good.)
So we have little to use for a study of voting behavior in American presidential elections.
Elections aren't the same, either. Some have incumbents running for re–election. Some don't. Some emphasize domestic issues; others stress foreign policy and national security. Probably the most useful — but far from infallible — approach is to compare circumstances. Is the election being held during a recession or boom times? Are we at peace or at war?
How will the historic nature of Clinton's candidacy play in all this? That is hard to say. If you want to compare this campaign to similar campaigns involving female nominees — Ferraro and Palin — you would conclude that Clinton will lose. But Ferraro and Palin weren't nominated for president; Clinton was. She is the first female to be nominated for president so there is no historical precedent to review.
Still, firsts
can succeed. Earlier in this post, I listed several historical firsts who lost, but Obama, of course, was elected and then re–elected. Yet, Obama was elected as much because of the economic implosion that happened just before the election as he was because of his status as the first black nominee of a major party.
I think a critical — and, although I have no professional experience in this area, I would think
immeasurable — factor is the often–mentioned
"fatigue" factor to which I alluded earlier. That makes it pretty tricky to be the candidate of the president's party. If the fatigue is as widespread as most election results have suggested it is after eight years, that candidate must run as the agent of both
change and
continuity. Only George H.W. Bush, who promised a
"kinder, gentler" version of the Reagan presidency in an attempt to appeal to centrists, succeeded.
Now it is Hillary's turn to persuade the voters that she can produce change while keeping things the same.
Recent history plays a key role here. Some states have been voting heavily for one party or another for several elections and, thus, are likely to continue doing so (although there is no guarantee; after all, realigning elections
do happen). States that have been narrowly voting for one party or the other, on the other hand, are more likely to
"flip" their allegiance. In modern America, this is most often seen in regions so let's examine the regions of America.
And, of course, turnout is a wild card. In poll after poll, majorities have had unfavorable opinions of both candidates. In past elections, most of those disgusted voters may have been considered likely voters. It is far from certain that these voters will be persuaded to support one of the nominees in this race. How will the outcomes in their states be affected if they choose to sit this one out?
In this study, I regard any state that gave no more than 53% of its vote to a candidate last time to be a prime prospect for flipping.
New England (Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, Vermont): Obama swept this region in 2012, and only New Hampshire gave him less than 53% of its vote. The other states voted heavily for Obama, which was not surprising since most of these states have been voting for Democratic nominees since Bill Clinton's first presidential election in 1992.
New Hampshire, with Republican roots that prevailed in spite of the presence on Democratic tickets of New Englanders John F. Kennedy in 1960 and Ed Muskie in 1968, has been an exception to the rule and could well be this time, too. At this point, I expect Clinton to carry the other five states and their 29 electoral votes, but Trump could carry New Hampshire's four electoral votes, especially since
he has endorsed Sen. Kelly Ayotte for re–election. Ayotte's support will be vital for Trump's hopes in the Granite State. She is the party's leading officeholder there, having been elected with 60% of the vote in 2010.
Polls truly are meaningless at this stage of the campaign. We need to get more distance from the conventions to get a good idea of where the campaigns stand, but recent history suggests that New England is Clinton's to lose.
Mid–Atlantic (D.C., Delaware, Maryland, New Jersey, New York, Pennsylvania): D.C. has voted for Democrats ever since it was first permitted to vote in presidential elections in 1964. The other states in the region have been pretty reliable for the Democrats as well, but Pennsylvania is usually a swing state, and recent history suggests Democrats have a lot of work to do there this time. Obama won more than 54% of Pennsylvania's vote when he ran the first time (the strongest showing by a Democrat in that state since 1964), but his share of the vote in that state dropped to less than 52% when he sought re–election in 2012.
There should be a lively contest for the state's 20 electoral votes. Meanwhile, Clinton currently appears likely to win the region's other 59.
South (Alabama, Arkansas, Florida, Georgia, Kentucky, Louisiana, Mississippi, North Carolina, South Carolina, Tennessee, Texas, Virginia, West Virginia): Most of the Southern states have been in the Republican column for the last 40 years. There
have been exceptions, though. In 2012, Obama narrowly carried Florida (with just under 50% of the vote compared to nearly 51% in 2008) and Virginia (with just over 51% of the vote; he snapped Republicans' 10–election winning streak in Virginia when he took more than 52% of the vote in 2008).
Obama carried North Carolina in 2008, but the state flipped back to the Republicans in 2012.
Clinton may benefit from having Virginia Sen. Tim Kaine on her ticket — at least in Virginia. The most recent Virginia poll I have seen was published before either of the conventions and showed the
candidates tied at 39–39.
Kaine hasn't won by historic margins — when he was elected governor in 2005, he received 52% of the vote, and when he was elected to the U.S. Senate in 2012, he received 53% of the vote — but they have been sufficient, considering that no national ticket from
either party has done better than that in Virginia since Bush–Quayle in 1988. If Kaine's supporters in Virginia turn out in November, Virginia will most likely be in the Democrats' column again.
As I say, most of the Southern states have been reliably Republican but a few could be primed for battleground status in coming years. Georgia, for example, gave Romney a little more than 53% of the vote, which was an increase over McCain's 52% four years earlier, which is considerably closer than presidential races have been in most Southern states, but not necessarily surprising in Georgia, which voted for Bill Clinton the first time he ran for president. Georgia is 30% black and 9% Hispanic. If those two demographic groups assert themselves, it wouldn't take much of the white vote to make elections truly competitive there.
South Carolina and Mississippi, with black populations of 28% and 37% respectively, could become competitive, but Republican margins have remained healthy there even with a black man topping the Democratic ticket.
Generally speaking, Trump should enjoy his greatest success on Election Night in the South.
Industrial Midwest (Illinois, Indiana, Michigan, Ohio, Wisconsin): This is the region where the election is most likely to be won or lost.
Based on electoral history, I think Illinois is really the only slam dunk for Democrats in this region. Illinois gave Obama (its former U.S. senator) more than 57% of the vote in 2012, but that was a decline from nearly 62% four years earlier. Still it suggests that Illinois' 20 electoral votes are secure for Clinton.
The rest of the Industrial Midwest, which has been hit hard by the economy of the last seven years, looks like it could be up for grabs.
Obama carried three of the other four states when he ran for re–election. Michigan gave Obama more than 54% of its vote in 2012 (down from more than 57% in 2008), but Michigan isn't far removed from the days when the vote was much closer.
Wisconsin, even with native son Paul Ryan running on the Republican ticket, gave Obama nearly 53% of its vote (down from more than 56% of its vote in 2008). Wisconsin has been voting for Democrats in the last seven national elections, but Obama was the first Democrat since Dukakis to seize a clear majority of the vote there.
Ohio is almost always a battleground state, and it is usually quite close. George H.W. Bush, in 1988, was the last candidate to receive more than 55% of Ohio's vote. Obama carried it twice, but his percentage there in 2012 was under 51%.
Historically, Indiana has been a lock for Republicans, but Obama's narrow victory there in 2008 leaves room for doubt. Obama trailed Romney there by more than 10 percentage points in 2012, which was more along the lines of what political observers have come to expect in Indiana. Trump's win in Indiana's primary all but locked up the Republican nomination, and I expect the state to be in the Republican column in November. The presence of Indiana's governor on the G.O.P. ticket can't hurt.
Midwest (Iowa, Kansas, Michigan, Minnesota, Missouri, North Dakota, Oklahoma, South Dakota): One electoral pattern that has emerged in recent decades is population–centric: Republicans have outperformed Democrats in mostly rural states while Democrats have outperformed Republicans in larger metropolitan states. That is a trend that has benefited Democrats in places like California, New York and Pennsylvania.
It has benefited Republicans in most of the states in the Midwest region. Romney carried six of the eight states in 2012, all but one with at least 57% of the vote. John McCain won those same six states but by smaller margins.
Missouri nearly gave Romney 54% of its vote but not quite so it barely qualifies as a state that
could flip to the Democrats. That could be something to watch on Election Night. In the elections that have been held since Missouri voted against President William McKinley's re–election in 1900, Missouri has only backed the losing candidate three times (Adlai Stevenson in 1956, McCain in 2008 and Romney in 2012).
Obama carried Iowa and Minnesota both times. Minnesota is probably the Democrats' most dedicated state. It has voted for Democrats in the last 10 elections. Richard Nixon in 1972 was the last Republican to carry Minnesota, but it hasn't been a slam dunk. Obama's share of the vote there in 2012 was about 52%, down from 54% in 2008. John Kerry took about 51% of Minnesota's vote in 2004, and Al Gore won with just under 48% of Minnesota's ballots (Ralph Nader drew more than 5% there that year).
In fact, Jimmy Carter's triumph there in 1976 with just under 55% is the Democrats' best showing in Minnesota since Lyndon Johnson took more than 63% of the vote in 1964.
Based on my 53% threshold for considering a state at risk for flipping, Minnesota should be on that list. But Minnesota has been consistent in its support for Democrats if not overwhelmingly so. Put an asterisk next to it. It
might flip — but it probably won't.
Iowa has only voted for two losing candidates in the last nine elections. Both were Democrats so that makes Iowa's record in that period 6–3 in favor of Democrats. And two of those Republican triumphs in Iowa came when Ronald Reagan topped the ticket in the 1980s. Iowa has been practically a regular in the Democrats' column for nearly 30 years.
But Democrats' share of Iowa's popular vote has been rather small. Less than 52% of Iowa voters endorsed Obama's bid for re–election in 2012. He got nearly 54% of Iowa's vote in 2008. Margins have been low on both sides. Republican Nixon, in 1972, was the last presidential candidate to receive more than 55% of Iowa's ballots.
Largely because of Obama's showing in Iowa four years ago, the state has to be considered a potential flip.
Rocky Mountain (Arizona, Colorado, Idaho, Montana, Nevada, New Mexico, Utah, Wyoming): The Rocky Mountain states were pretty reliable for Republicans until recently.
Arizona, with a Latino/Indian/black population that makes up two–fifths of the state's overall population, has voted for Republicans in the last four elections but not by the kind of margins that were seen there as recently as the 1980s. Bill Clinton won the state when he sought re–election and narrowly lost it four years earlier. Since the start of the new millennium, only George W. Bush in 2004 has received more than 54% of the vote there. Arizona could possibly flip in November.
Colorado has been trending Democrat in recent elections, but it gave Obama less than 52% of its vote in 2012. It
could flip to the Republicans. What happens there on Election Night will be worth watching.
Idaho routinely gives Republicans more than 60% of its vote. So do Utah and Wyoming.
Montana is less reliably Republican but still gave Romney more than 55% of the vote in 2012. McCain barely won the state in 2008, and Montana was close in the 1990s so it might bear watching on Election Night.
Nevada is a modern bellwether, having voted for every winning candidate but one (Gerald Ford in 1976) since 1912. It gave Obama just over 52% of its vote in 2012, which makes it a possible flip.
New Mexico has been supporting Democrats by and large in the last six elections. It gave Obama just about 53% of its vote in 2012, which makes it a possible flip, but with a Latino presence that represents more than 46% of the state's population and another 10% of Indian and black backgrounds, a flip seems highly unlikely.
Pacific Coast (Alaska, California, Hawaii, Oregon, Washington): Since Alaska became a state in 1959, it has voted Republican in every election but one. Incumbent Republicans (with the exception of George H.W. Bush in 1992) generally do better than nonincumbent Republicans, which suggests that Trump may not win Alaska decisively, but he is still likely to win it. Four years ago, more than 54% of Alaska voters voted for Romney, which was down from more than 59% for the McCain-Palin ticket four years earlier. Of course, Palin was Alaska's governor at the time.
Clinton is just about certain to win California and Hawaii by wide margins. California has only recently been giving lopsided majorities to Democrats. Obama got more than 60% of the vote in that state both times, but prior to that no candidate had received more than 60% of California's vote since Franklin D. Roosevelt in 1936. Not even Californians Richard Nixon and Ronald Reagan.
Still, even if Clinton is held to a more historically typical share of about 55% of California's vote, she would probably win the state by about 2 million votes. Minority groups combine for well over 50% of the electorate in California, and you don't need a poll to know Trump is not popular among minority groups.
Hawaii gave Obama more than 70% of its vote in 2012, which was down a point or two from four years earlier — but that may well have been a byproduct of Obama's Hawaiian roots. The only other presidential candidate to exceed 70% in Hawaii was Lyndon Johnson in the landslide year of 1964. Otherwise, even though Democrats usually win Hawaii, the winning percentage has tended to be in the low to mid–50s.
It wasn't terribly long ago that Oregon and Washington voted for Republican nominees regularly. In more recent elections, both states have trended Democratic, and their percentages from 2012 suggest they are not likely to flip.
Thus, Clinton is likely to win all the Pacific Coast states except Alaska.
Conclusion: Clinton is the likely winner as of Labor Day, but there are still more than two months to go.
