Donald Trump’s campaign strategy was ridiculed from here to Timbuktu by political nerds, GOP and Democratic strategists and the media. Trump staked his claim to victory on riding a wave of populist sentiment to victory across the Midwest.
By any standard definition, Wisconsin would not be thought of as a tipping point state (see definition here) but it was. However, so confident was the Clinton campaign that Trump was not competitive in the state that they visited the state, count it with me now, zero times. She did not set foot in the state after losing to Bernie Sanders in the Democratic primary in April.
To be fair, Wisconsin and other Midwestern Blue Wall states had demographics that did not favor the party. They are chalk full of blue-collar whites, Trump’s bread and butter, polls showed Clinton ahead, and Sunbelt states were looking due to demographics and polling to be in reach.
Of course, this did not turn out. Clinton lost every Blue Wall state except Minnesota. Obviously, you could say that Clinton should have visited Wisconsin. She should have gone to Michigan more than once and she should have done more to appeal to rural Pennsylvania. But, that said, this probably did not cost her by itself the election.
No, what cost Clinton the election was Trump’s gamble on stealing Democratic electoral votes. Trump gambled due to demographics, the industries of these states and their reddening shift since 2014 would allow him to eke out victories over a weak Clinton. Meanwhile, Clinton staked her campaign’s victory on winning states like North Carolina and Georgia. Who had the better strategy is in the winner of the tipping state.
Typifying the states the candidates considered most important can probably be gleaned by simply looking at where they spent their time. By this measure, both Trump and Clinton considered Florida to be the most important state, followed by North Carolina, Pennsylvania and Ohio for Clinton. Trump considered Pennsylvania the second most important state followed by North Carolina and Ohio. But after that the visits look completely different with Trump spending time in Colorado, Michigan and Wisconsin while Clinton spent time in Iowa, Nevada and Georgia.
If one digs deeply they can find some of these visiting differences are found on the margins. Rather, it seems Clinton made two significant tactical mistakes (at least according to fivethirtyeight and ones I agree with).
Clinton focused only on close states: According to RCP averages the closest states were Florida, Iowa, Ohio, Nevada and North Carolina. Obviously the polls were off in Wisconsin, Michigan and Pennsylvania but this was not reflected until the final results.
Where Clinton made her mistake was focusing only on these “must-win states.” If Clinton won these states she would hit around 300 electorate votes. That is nothing to sneeze at but if you lose Iowa and Florida you lose 35 electoral votes and the race is a toss-up. Clinton never tried to cushion her margin by focusing on keeping already blue states in her column. If the campaign had paid attention they would have notice the shifting winds late in the campaign and seen Trump gaining strength among the blue-collar workers who supported Obama a mere four years earlier.
Fivethirtyeight notes that Michigan, Wisconsin and Pennsylvania up to the beginning of November narrowly mimicked the national polling averages (as in they were most likely to be the tipping state). Obviously, they did not follow the trendline in the polls after the beginning of November but they did in the final results. So, if the race narrowed, as it did before Election Day it would follow that these states would also be more competitive (regardless of what the polls said).
Even if you give Clinton the benefit of the argument and argue a good offense is a good defense Clinton’s campaign blundered. Clinton only made a token effort to steal GOP states like Georgia and Arizona whereas Trump ran all over the place at the end of the campaign. He even went to Minnesota. Whereas Florida and North Carolina were not necessarily must win states for Clinton they were for Trump. Clinton, instead of focusing on Ohio and Iowa, states that were not near the national polling average, should have parked herself in these states and focused on not letting Trump gain even a smidgen of momentum.
Clinton’s overconfidence showed in the narrow range of states she campaigned in: Clinton played to a much narrower base of states to win than Trump. Where Clinton spent no time in Virginia, Minnesota, Colorado or Minnesota, Trump spent time in all of them. He actually won none of these states but the point is that he was trying to expand the map.
The critical difference here between Clinton and Trump is that Trump’s visits reflected the overall uncertainty of the race while the Clinton camp assumed the race was all but over. Indeed, the massive polling shifts in the race we saw time and time again should have been an indication of the volatility in the electorate.
The fact there were indications of volatility in the race could have led to any number of things happening; suburbanites could have swung to Trump, turnout in urban areas could have surged, etc. The point is that the polls, the media and all the analysis in the world can be wrong, and because these things lead to unpredictably you want to make your path to victory as wide as possible.
Supposedly, the Clinton campaign had the most data savvy team ever assembled. So why did the Clinton camp make this mistake? Because they bought their own hype. Internal models were incredibly overconfident and alternate models built on lower urban turnout, lower college educated turnout, never occurred.
This problem can be compounded with internal polls. Campaigns only spit out good internal polls. In states like Wisconsin, Pennsylvania and Indiana this is why few GOP candidates put out internal numbers but Democrats spit them out up and down the ballot.
Of course, where to spend money and a candidate’s time is never made in a vaccuum. Internal politics and external politics have an impact. In Clinton’s case this was certainly true. Clinton’s campaign was so intertwined with the DNC that millions were spent to run up the vote in big cities in safe red Kansas City) and blue states (Chicago). Clinton’s campaign was religiously devoted to control. As a result, the campaign did not provide local, on the ground volunteers. Finally, the campaign did not run a GoTV campaign until the final days of the campaign assuming their turnout would not need to be substantially goosed.
If you are confused by this you would be forgiven. If you thought the Clinton campaign had a robust, well oiled volunteer machine you would be forgiven. It was entirely driven by the media and perception.
Even worse, coverage of the relative strengths and weaknesses of the Electoral College was missing any context. For example, on Nov. 3 the New York Times went after Trump for campaigning in too wide a range of states (including Michigan, Wisconsin and Pennsylvania). This too wide range of states is exactly the reason Trump won. With the race tightening Trump found the soft spots in Clinton’s weaknesses in these states. The rest is history.
Speaking of context, on the ground analysis was often missing or half-assed (pardon my french). For example, if Clinton spent money in Indiana or Arizona to boost downballot Democratic candidates it was a sign of strength. But, if Trump showed up in New Mexico or Minnesota it showed just how desperate he was. Indeed, this was just a continuation of the press fueling the narrative Trump did not have a coherent campaign strategy.
Speaking of the media, it was clear just how much it drove the election. Whenever Trump was in the news he sank in the polls. Whenever Clinton was in the news she sunk. To a degree this might explain why Clinton never fully abandoned states like Ohio and Iowa (where she consistently polled behind) which would have led to media freakout. It also helps explain why the campaign made questionable tactical decisions to spend money in Indiana and Arizona. It generated positive headlines.
Credit needs to go where credit is due though. Trump’s data analytics team, despite getting little credit, should get some. They saw where Trump was weak and kept him away from those states (he did not show up in CA or NY as he said he would earlier in the campaign). In turn, they honestly assessed where he could win, how and directed his attention to those places. The media would never give that much credit to this successful strategy and has not.
Indeed, short of a single interview, Trump’s analytics team at Cambridge Analytics has largely been obscured until recently. Even the models they built saw Trump winning very, very rarely. But, the proof is in the final results. And so is the vindication of a campaign strategy derided for so much of the campaign.