Talk Elections

This one puzzled me because it was the first state, along with Vermont, to be called for Clinton on election night.  But it ended up being the closest state that year, with Clinton winning by 0.59%.  What happened?  Did the TV networks rely too heavily on flawed exit polling, like they did in 2000 when they prematurely called Florida for Gore?

Yeah, Clinton ended up winning NH by just 1 point (huge shift from 1988 when Bush won it by 20 points; his second best state after Utah).  The liberal media is so dishonest that it's ridiculous.  Bush would have comfortably won Florida in 2000 if they did not prematurely call it for Gore.  The heavily republican panhandle is in central time zone, and lot of Bush voters stayed home when they heard that Gore had already won it.

Care to explain how Democratic media bias played a role in networks calling Mississippi and Alabama for Reagan when polls closed? Even though he won both by only 1.3%? Maybe exit polls in the South just suck(ed).

Clinton actually underperformed his polling expectations; he was expected to win by double digits, so networks thought it would be a pretty safe bet to call states that were most likely gonna go to him. I bet they thought Oklahoma would vote similarly to Arkansas, which wouldn't have been a terrible assumption at the time. There was also a big Perot effect that probably messed up exit polls.

Besides, after the 2000 debacle, networks got a lot less lenient on how soon to call states.

In 1980, some states were early called to Reagan, and his margin was very small.

In 1992, some states that Clinon won narrowly were called early to Clinton, and some states that Bush won were considered too close to call.

In 2004, some states that Kerry won narrowly were called early to Kerry, and some states that Bush Jr. won narrowly were considered too close to call.


Hypothesis:
Exit polls do not reach well the rural areas of the states. So, they overestimated Clinton and Kerry, and overestimated Reagan too. In the southern states, Carter did better in the rural areas.

Why would exit polls be less reliable in rural areas?  From my understanding of how exit polls work, people ask random voters as they are leaving the polling station, who they voted for.  Would rural voters be more inclined to lie than say urban or suburban voters? Or is it simply a numbers issue whereby there are fewer volunteers in rural areas and hence the sample size is too small to make a reliable conclusion?

In 1996 Clinton was leading Dole by as much as 15 points, but in the last 2 weeks of the campaign Clinton basically stopped campaigning for himself and instead focused on congressional races.  Perhaps the late polling that year didn't quite capture that effect.

Exit pollsters are less likely to sample rural precincts. 

The call was made about 10 minutes before the polls in the Central Time Zone closed....

It's important to note that Vermont voted straight Whig before the Republican Party even existed. Why was it Republican for so long?

The question is "why is it so Democratic now?"

-Because it was so strongly Republican before 1960, and many older Republicans in it couldn't believe what the GOP was becoming for decades.

If Georgia didn't go straight to Bill Clinton at 7 PM, Georgia would've been the only one too close to call by 3 AM, Clinton would have to wait a week or two to win the Peach State.

A total of 300 EVs went to Bill Clinton as soon as the polls closed in 1992, and in 1996, a total of 350 EVs went to Bill Clinton as soon as the polls closed in those states.

In 1992, Bill Clinton carried my home state of NJ by a margin of just 2.37%! (In 1988, Bush won it by 14 points, and is the only state excluded from the Dukakis strategy to not cast a single vote for a Republican since the 1980s.)

Maine was excluded from the Dukakis strategy too. So were NM, NH, VA, CO, and NV!

This is also the only election where Vermont and Georgia both voted Democrat in the same election (if you don't include the elections before the founding of the Republican party). Vermont never went Democrat until 1964, Georgia never went Republican until 1964. Both states voted Republican in 1972, 1984, and 1988.

Wait a minute, Delaware was also excluded from the Dukakis strategy, and it hasn't gone Republican since 1988. Surprise, right?

If I was Bush in 1992 or Dole in 1996, I wouldn't even set up a winning ceremony.

Republicans were the party of Protestant Yankees from birth. Whigs were as well but not the same extent. A lot of free soil Democrats in VT never went back after joining the GOP. There was a Jeffersonian/Jacksonian element within the state, but it got largely absorbed as well.

Politics became a generational thing and VT had a very strictly adhered to political establishment within the GOP and those that tried to skip in line, were often destroyed politically. That mattered more than ideology so there were a large number of progressives and "traditionalist" (Protectionist, anti-immigration, isolationist) Conservatives within the same party in the state.

In the 1950's and 1960's, Conservatism became defined nationally in a way that made it hostile to the state's voters, especially on defense issues. The Conservative wing collapsed and the first Democrats got elected in 1958. After that, the only ones who could win were the heirs to the Progressive wing (Prouty, Aiken, Stafford, Jeffords), and as the state's population exploded from New York transplants,  the Protestant religious fervor pretty much collapsed, and the rise of environmentalism became a factor.

By 1992, enough old people who still were part of the tribalistic Republicanism had passed away, leaving the state open to voting for a Democrat. And it hasn't looked back since.

I think a very underrated factor in Vermont and New England's political transformation is that the region was the heartland of an ancient century religious left movement (not exactly leftist in today's terms, but on the issues that carry over from that time) that collapsed almost overnight after they finally got a child labor ban past SCOTUS in 1938.  That was a 100 year struggle from the mid-19th to the mid-20th century, and after it was finally won, the movement was divided on war and peace and never really found a new issue to coalesce around.  Prohibition and its ultimate repeal also discredited them significantly.  It wasn't a clean R vs. D split, but more a Protestant/Catholic divide with both parties having a foothold in the movement and in the establishment that resisted it.  Vermont was on the establishment side of this divide after 1890 or so and steadfastly opposed FDR.   

Ironically, I would imagine the modern religious right would find itself in a similar crisis if Roe v. Wade ever truly goes down.

CBS's list of battleground states in 2016:

Georgia - 5.1%
Virginia - 5.3%
North Carolina - 3.7%
Ohio - 8.1%
New Hampshire - 0.3%
Pennsylvania - 0.7%
Florida - 1.2%
Arizona - 3.6%
Colorado - 4.9%
Wisconsin - 0.8%
Michigan - 0.2% (went from the bottom of the close state list in 2012 to the top in 2016)
Iowa - 9.4%
Nevada - 2.4%

Minnesota isn't even on there... It was a 1.5% state, Maine was a 3% state.

Also, news networks prematurely called Pennsylvania for Barack Obama in 2012. If they didn't, Romney probably would've won it. (sure, by about the same margin as in North Carolina)

NBC's staff said "they need to win Pennsylvania, the election is over if the Democrats lose Pennsylvania." If Obama lost Pennsylvania, he still would've won (312 to 226).

The Georgia senate race was too close to call.

The polls were way off both in 1992 and 1996 in Clinton's favor.  In 1992, the October polls had Clinton up by 10-15%, him only winning by 5.5% was a surprise and underachievement. In 1996, Clinton was up 15%-20% in the polls leading up the election and only won by 9%. The media was expecting a big Clinton blowout both times, in the mold of Reagan 84 or Eisenhower 52/56. These terrible polls had a partial effect in the media calling those states early for Clinton.

I'm surprised Delaware was too close to call in 1992.

In 1996, only four states that were undecided went to Clinton later on: Kentucky, Louisiana, Nevada, and Arizona. (Louisiana would go for Clinton by DOUBLE DIGITS!)

Georgia has a law that requires the winner of a GE race (including Presidential electors) to achieve 50% to avoid a runoff.  Georgia was projected for Clinton early, but taken off the board later on after reporters learned that there was a runoff  (Fowler was close to 50%, but Clinton won Georgia with 43% as Bush refused a runoff.) 

Maybe the networks started being more cautious after Georgia 1992, and even more cautious after Florida 2000. Since 1976, every time Georgia was undecided, it went Republican. In 2004, Georgia went Republican on poll closing time. They didn't call states at poll closing time until 1980.

Anyway, in Georgia in 1996, NBC said Dole was 1 point ahead all night. But in 2016, Trump started off with an immense 50+ point lead before collapsing to a 5 point lead. They projected the state when Trump's lead was much smaller. That isn't what they did in 2012 with Pennsylvania. When Fox News projected the state would go to Obama, he was leading by 33 points with 7% of the vote in, this was ultimately a 5 point lead.

Dole won Georgia with 47% of the vote, so Clinton must've refused the runoff. Is that correct?

If Clinton won Georgia in a runoff, that would give him 32+DC again, he would have 392 electoral votes. Without Georgia, he already had 379, which was enough for him to win.

These people sure like to give Florida's 25 to the Democrats. It went blue once and red twice. In 1992, some news networks incorrectly projected Clinton the winner, and in 2000, it's obvious, Gore was incorrectly projected the winner. In 1996, when it went blue, the state was the first state Clinton won on election night.

I assume the huge shift in New Hampshire was due to the projections from news networks. This is another state they like to give to the Democrats. In 1992, it was projected at poll closing time, it was won by 1.1%. In 1996, they falsely projected at poll closing time Dick Swett would defeat Bob Smith.

Correct me if I'm wrong, but didn't they also used to call states before their polls even closed?

Maybe. Not since 2004 however. Only one certain case--the state of Florida for Al Gore in 2000.

They handed Michigan's 18 electoral votes to the Democrats at 8:00 three times.

In 1988, they projected a very close state for Dukakis at poll closing time, it was New York.

Georgia has never had any such law applying to presidential elections.

Fun fact: when Georgia originally adopted its runoff law, it actually did apply to Presidential elections! This remained the case until the law was changed in March of 1968 to exempt them from the runoff requirement (otherwise a runoff would have been required after that year's election. This wasn't a coincidence - Wallace supporters in the General Assembly were expecting a win and didn't want a runoff to jeopardize it)

Now that is news to me! Do you know the year that runoffs were extended to the general election? For primaries it dates back to 1915 but of course they were also using County Unit until 1963; and the 1966 governor's race had to be decided by the General Assembly not through a runoff.

I've found the 1968 law in question and it appears the original act which is amended took effect in 1965 meaning the runoff provision was never actually applicable to a presidential race. But what a fascinating tidbit of history.

http://metis.galib.uga.edu/ssp/cgi-bin/legis-idx.pl?sessionid=4948683c-01c50f8e43-0428&type=law&byte=295414869

As to the 1992 point, again, there was no possibility of a runoff for presidential electors.

Here is what happened:

In the 1996 NH senate race, The Union Leader said the exit polls had a measly 5% chance of being off by 3%, there were three polls that had Swett ahead by about 5%. (Smith however won.)

In 1992 in Georgia, Clinton was ahead in a poll by 6 points.

Had this persisted in 2016, Wisconsin would've been called for Hillary Clinton at poll closing time, even though Trump ultimately carried the state.

Alabama in 2016 was undecided, Trump won that state by... TWENTY-EIGHT points.

This was more theoretical than mathematical. PA had consistently been polling a couple points to the left of the nation prior to the election, so a Romney win in PA would have indicated that he was winning the NPV by 3-4 points.