WikiLeaks Didn’t Cost Clinton the Presidency in 2016: Data
This is the first in a series of posts challenging the Trump-Russian collusion conspiracy coup that is being forwarded by many special interests and the press that aren’t happy with the 2016 election results. This post tackles the so-called “interference in the election” concept and whether or not hacking, by the Russian government, Russians, or others, that was later shared with and released by WikiLeaks, caused Democrat Hillary Clinton the presidency in 2016.
The simple answer is that, No, WikiLeaks didn’t cost Clinton the presidency.
Pre-vote polling from swing states that RealClearPolitics.com was tracking – hundreds of them – shows that in the polling averages in each state, Clinton received a higher percentage of the votes than she was expected to receive everywhere that she lost. In every swing state where Clinton won, she also received a higher percentage than the polling average in each state. There were some state polls that showed Clinton receiving a slightly higher poll position than the votes that she received on Election Day. However, they were not only rare but within the standard margin of error. And, lastly, in only three pre-WikiLeaks release polls in states Clinton lost – out of 32, in other words, less than 10 percent – was she in a higher poll position than the actual results. However, all of those poll results, too, were within the standard margin of error.
Hacking timeline
Before eyeing the data, it’s important to understand the entire Trump-Russia conspiracy coup timeline, to get an understanding of how the polling data fits into the WikiLeaks release.
Both the RNC and the DNC were warned that some entity – later, accused to be the Russians or hackers funded by the Russians although this has still not been proven; it’s speculation – was trying to hack into the orgs’ computer systems in August 2015. The RNC took precautions; the DNC ignored the phone calls, according to press reports.
In March 2016, Clinton campaign chairman John Podesta – who was using the word “password” as his gmail account password – clicked on a phishing scam in his inbox after an aide sent out an email, with a typo in it, to staffers telling them to change their passwords. The staffer accidentally told Podesta the scam email was legitimate when he meant to write that it was illegitimate.
A month later, in April, the DNC discovers what appears to be two hacks on its networks.
Trump secures the Republican nomination by vote count in late May (although, the entire situation is unstable until the actual convention begins in mid-July).
Bernie Sanders concedes to Clinton on July 11.
WikiLeaks released the first batch of emails on July 22, 2016, right before the Democrat’s convention on July 25, 2016, and released subsequent batches of information before the election.
I won’t get into all of the specifics of what was in the WikiLeaks releases beyond stating that they showed that the Democrat’s 2016 primaries were actually rigged against Sanders and there was rampant media collusion with the Clinton campaign, too. No one knew any of this was going on, since it was secret.
Why look at polling data?
While polling data isn’t always completely accurate, it is a private snapshot into what a sampling of voters are feeling at any given time. In other words, it is often all anyone has to go on. Presumptions can be made from there and they are often spot-on correct. Since state polls are more local than national polls, the information in them is a better reflection than say a national poll from all around the country.
In 2016, according to RCP, there were 16 “battleground states,” probably the most for any presidential election in the modern era as we know it (it tends to come down to less than 10). Many of these – despite the reporting – were never really swing states. But because, here and there, there were polling anomalies and the election was so volatile, more states were tracked. This occurred, most analysts presume, because voters were not telling pollsters that they were voting for Trump, for any number of reasons, including fury at the media, fear of backlash for sharing their choice with a pollster, fear of targeting later since, often, polling is now doing by political orgs in order to peel off votes from the other side, etc. Nearly everyone in the political world and the polling industry knew that Trump’s polling numbers were depressed.
Swing states Clinton lost
The data below, in two sections – swing states Clinton lost and those she won – shows the pre-vote polling collected by RCP, the peak poll results for Clinton, her average in each state, and then, those numbers were compared with the actual vote results. From there, the pre-WikiLeaks release polls were eyed and compared those to the end results for Clinton.
The data shows that in every state that Clinton lost, she received a higher percentage of actual votes than the polling average predicted she would receive. In other words, WikiLeaks didn’t suppress Clinton’s vote – she got more votes than predicted by the polling.
Here is the data from the swing states where Clinton lost:
Arizona: AZ was listed as a swing state but never really has been in the past – it’s been solid Republican for a long time. Some polling, however, in mid-to-late August and again in early October, had Clinton leading by 1 to 5 points, garnering – at her highest point – 45% in mid-October. Trump had 46% in that poll. Clinton’s numbers ranged from 38 and 45 during the final weeks of the campaign while Trump ranged from 38 and 49.
The RCP average through Nov. 1, the last poll taken, was 46.3% Trump and 42.3% Clinton.
The election result was Trump 49.5% and Clinton 45.4%.
Clinton received more votes than the polling average said she would ever receive, since she never broke 45% in any poll.
RCP listed no polls from the state before the CBS News/YouGov poll from Aug. 2-5, so it’s unknown what Clinton’s pre-WikiLeaks numbers were in the state.
Florida: This state has been a seesaw battle for generations between the two major parties and in 2016, it was no different. While Clinton had a couple of leads in FL polls in June and one in mid-August, most of the polling was within the traditional margin of error – 3% to 5% – with both candidates trading the lead, back and forth. In only eight out of nearly 70 polls did Clinton get as high as 48 or 49. In most, she was in the low-to-mid-40s.
The RCP average had Clinton receiving 46.4% and Trump receiving 46.6%.
Trump won the state by 1.3% with Clinton receiving 47.8% of the vote.
Clinton’s result was better than expected based on the averaging and nearly 90 percent of all polling in the state.
There were seven polls released between June 2, and July 11, before WikiLeaks, with both candidates shifting leads. Clinton’s highest registration was in the last poll – 47%, with an average of 40.5%. So, her actual vote result was also better than before WikiLeaks docs were released.
Georgia: GA, like AZ, was a swing state this year but never historically has been except when southern Democrats, like Bill Clinton in 1992, are on the ticket (although, Bob Dole won the state in 1996 by less than 30,000, with H. Ross Perot siphoning off potential Republican support). Hillary Clinton had four shock polls in late July, and early-to-mid-August that showed her either tied or up to 7 percent, something no one in the political world truly believed even if the “Georgia is in play” made for good headlines. In every other poll, Trump led, and in three, he got to – or cracked – 50%.
The RCP average had Trump by 4.8% – 49.2 to 44.4.
Trump won the state, 51.3 to 45.6.
Clinton, again, received a higher percentage than she was expected to receive based on the polling average.
GA had a single poll before WikiLeaks and Trump led that poll by 7 points. Clinton was polling at 38 percent – more than 8 points less than what she received on Election Day. WikiLeaks didn’t depress her vote there.
Iowa: IA has, consistently, been in play since the 1980s, and has gone to the Democrats in five of the last eight presidential elections. In 2016, Trump won in a landslide – 9.6%. With the exception of a single poll in late June that had Clinton up by 13%, most of the polling had the numbers either within the margin of error or Trump up by as much as 8 points.
The RCP average had Trump winning by 3 points, with Clinton getting 41.3%. She ended up with 42.2% of the vote or more than expected when compared to the average.
Only three of the 18 polls had Clinton above 42% (all three were 44%; two showed Clinton ahead of Trump and one was a tie).
Five polls were tracked by RCP before WikiLeaks and in three of those polls, Trump led by tiny margins or it was a tie. Clinton led in one poll by 2 and another one – a June 24-28 Loras College poll – by 13 points. That poll, one of the three that had Clinton at 44%, recorded Trump with 31%. Her RCP average for the five polls before WikiLeaks was 40.2%.
If the results were tighter, one could presume that WikiLeaks may have depressed Clinton’s vote, and it might be worth eyeing local media coverage, like newspapers, television news, and talk radio, to see if any of them spent any significant amount of time reporting on or discussing about WikiLeaks. Trump’s landslide result negates this though.
Michigan: MI was one example that was the opposite of AZ and GA in that Clinton led in all but one of the state polls, which started in early July (that poll was the very last poll of 1,200 likely voters released the day of the election, put together by the Trafalgar Group, a GOP pollster, and had Trump up by 2). In fact, she led by as much as 13% in one poll and 11% in four polls. However, in only three polls did she get to 50 or 51%; in most polls, she was in the low-to-mid-40s.
Her RCP average was 45.4% to Trump’s 42%.
Sidebar: During the early morning hours of Election Day, Trump held a boisterous rally in front of thousands of people in Michigan and I was watching it on C-SPAN. I remember thinking at the time that this was a mistake because any good campaign wants its supporters and worker bees rested up in order to get-out-the-vote (GOTV). But Trump’s campaign, as we all know, was not a traditional one, and utilized as much free media as possible (since it was being outspent by billions). This rally turned out to be a worthy gamble, it would seem, since Trump won the state by less than 12,000 votes – or 0.3%.
Clinton’s 47.3% result, however, was nearly 2% better than her RCP average, with Trump receiving more votes than expected, even above the margin of error. In only five of 32 polls though was Clinton above her actual results.
One could make the assumption that with such a razor thin margin that any negative information may have depressed Clinton’s or boosted Trump’s turnout. However, the juxtaposition of that is she received a higher than the average vote tally and was within the MOE of her highest total in any of the polls. As well, Trump was under polled, as we already know.
Turnout was also higher in 2016 than in 2012 – about 73,000 more people participated last year than the previous four years. And, as noted in this previous post, Trump received more than 185,000 more votes than Romney in 2012, a state where his dad was once the governor and another relative runs the Republican Party.
In pre-WikiLeaks polling – there were three – Clinton’s highest result was 42% with an average of 37.67%, or lower than most of her post-WikiLeaks polling.
Missouri: Like the other southern “swing states,” this one was never in play but a shock poll from the St. Louis Post-Dispatch turned up in late July showing Clinton up by 1 point, which shoved it into the toss up category. Like GA, history shows that MO is never in play for Democrats unless there is a southern Democrat on the ticket like Bill Clinton – who won it in 1996 and 1992 – and Carter, who won in 1976.
The RCP average had Clinton at 39.3% and Trump winning by 11; Trump won by 19.1%, with Clinton receiving 38.
This is another example where Clinton’s actual outcome was less than the average albeit within the margin of error.
Did WikiLeaks hurt her here? Nope. The shock poll was put together a day after the first release of documents. As well, there was only one other poll by PPP, a Democrat firm, nearly two weeks before WikiLeaks, and that poll had Trump up by 10.
North Carolina: NC, which historically has been safe for Republicans, has become a swing state in recent years (Barack Obama won the state in 2008 … the first Democrat since Carter won it in 1976). Up until mid-to-late September, Clinton and Trump traded leads in the state with Clinton as high as 9 points in early August and Trump as high as 5 points in late September. Between late September and late October, Clinton led by single digits in every poll but five of 27 polls (two Republican firms showed Trump with a 3- to 5-point leads while two media outlets showed Trump in a tie and up by 7 percent). The last week of the election, three polls were released showing a tie and Clinton up by as much as 2.
The RCP average though showed Trump winning, 46.5 to 45.5.
On Election Day, Trump won by nearly 4-points, 50.5% to 46.7%.
Clinton, again, received a better than the average poll vote outcome: 46.7%, although in 11 of the 51 polls, in other words, about 20%, she received higher than her actual results, 47-48%, while getting to 49% twice.
In NC, there were six polls before WikiLeaks: Trump led two, one was a tie, and Clinton led two. In one of Clinton’s lead polls, she was at 49%, with a 9-point lead (Civitas, a Republican firm). But this poll was in late April, when she wasn’t even the nominee yet. Her RCP average for these six polls was 43.17%.
In other words, WikiLeaks didn’t influence the result here either.
Ohio: OH has also been a swing state for a bit with both parties eking out presidential wins. Polls tracked since early June showed leads shifting back and forth with solidifying leads late in the race. By mid-October, Trump led or tied Clinton in every poll, racking up 5 to 7 point spreads in four polls in the final weeks and eight polls in the final two months. Trump campaigned in the state hard on the issues of coal and energy and he used bad trade deals ushered in by Clinton’s husband against her – and the strategy worked. It wasn’t even close: Another Trump landslide, by more than 460,000 votes, nearly 9%, with Trump receiving more votes in the state than either Romney or Obama in 2012.
The RCP average for Clinton was 42.3% but she received 43.5% on Election Day or, again, better than the average.
RCP tracked 35 polls from early June to early November. In less than half – 14 – Clinton did better than the votes she received (44%-46%); in half of those polls though, Trump had the lead or was tied with Clinton. Only two of the polls – in mid-August and early September – were outside of the margin of error.
There were five pre-WikiLeaks polls; Clinton led in four, within the MOE, with one – CBSNews/YouGov from July 13-15 – polled her at 44 percent, 0.5% higher than her actual result. The RCP average of the five polls was 39.8%, or less than the end result.
Again, no WikiLeaks influence here either.
Pennsylvania: Not unlike OH, PA was also a place where Trump campaigned hard on trade and coal. But with the exception of a single poll in late June – which had Trump up by 6 – and two GOP polling firms that had Trump either tied or up by 1 point in the waning days of the campaign, Clinton led in every other poll – more than 40 of them – by anywhere from 1 to 11 points; 18 were outside the margin of error; 10 of those were before October. In other words, the race was tight there, too, as expected. Clinton, however, only polled as high as 50 in a single poll in late September.
The RCP average had her at 46.2 with Trump coming in at 44.3.
Voters stunned everyone though, casting more than 2.9 million ballots for Trump and he won by nearly 70,000 votes and, again, received more votes in the state than Obama or Romney in 2012. Clinton, again, received more than the average expected vote though: 47.6%. Less than a third of the polls had her at 48 or above.
There were four pre-WikiLeaks polls and Clinton led in three of the four by 1 to 8 points (the other was that shock Trump poll from June showing him with a 6-point lead). The highest Clinton polled in these four polls was 43 and her average was 39.25%, all below her actual results.
While PA was a shocking win for Trump like MI, there appears to be no negative WikiLeaks influence on Clinton’s result. This state was probably the most under-polled for Trump in the election.
Wisconsin: WI is another traditionally Democrat state that has become a swing state (Republicans have only won three times in the last 10 presidential elections). Clinton led every poll tracked between late June and the election in the state. One poll in August showed her with a 13-point lead but she never polled any higher than 49% and that was once, the week before the election.
Her RCP average was 46.8 to Trump’s 40.3. The end result was a Trump win: 47.9% to 46.9%.
Of all of the swing state poll averages, this one was the closest predictor for Clinton: 0.1%. But, again, it was better than expected. In five of the 19 polls, Clinton polled at 47 to 49, or higher than the results, again, about 20 percent.
There were two pre-WikiLeaks polls showing Clinton with leads at 43 and 41, or below her actual results.
Sidebar: Frankly, I was stunned that Trump won Wisconsin and the spread between his 40.3% polling average and 47.9% result is huge. At the same time, while turnout was a little more than 3% lower in 2016 when compared to 2012, Trump received more votes in the state than Romney (less than 1,000). Clinton, however, received 213,000 fewer votes than Obama (even though she received nearly exactly what the polling said she would receive). This is why she lost the state: Clinton could not hold the Obama base. In places like Wisconsin, they went to Trump, voted for indies, or stayed home.
Data conclusions on states Clinton lost
The only conclusion anyone can come to after looking at all of the evidence here in the polling data is pretty overwhelming: WikiLeaks had absolutely no influence on Clinton losing the 2016 presidential election. In every state, her pre-vote polling average was lower than her actual vote result.
WikiLeaks has also continually and consistently denied that the release information came from the Russians. But it doesn’t really matter at this point, does it? The release of the hacked information didn’t interfere with the election beyond politico fodder.
Take this one step further … and this is important because the entire Trump-Russian collusion conspiracy coup is tied into this … the Russians were supposedly colluding with Trump (who was not the nominee in August 2015) to influence the election by hacking or paying hackers to get at the DNC (and don’t forget, they tried to hack the RNC, too, even though the press is under mentioning this), sending a phishing scam to Podesta in March 2016 (when Trump wasn’t even the nominee yet either), in order to hurt Clinton (who also wasn’t the nominee yet either), and the hacking uncovered a treasure trove of actual collusion in the Democrat’s primaries (countless examples, but the hackers didn’t know this would actual result in any real information that could harm Clinton who wasn’t the nominee yet), the hackers give the information to WikiLeaks to release (on the eve of Clinton’s nomination), which WikiLeaks releases (but it’s under-reported), in some grand conspiracy in order to keep Clinton from becoming president … phew … but, but, but … the data shows that the release of the information had no negative influence on Clinton’s results in any way, shape, or form, in any swing state, according to the data I’ve posted above … combine this data with the revelation in the book “Shattered” that Podesta and Robby Mook cooked up the “Clinton’s Russia-blame-game” conspiracy the day after she lost the presidency … and the Trump-Russian collusion conspiracy coup ends right now!
Swing states Clinton won
Here is the breakdown of the swing states that Clinton won. I’m including them because it shows consistency with the polling in the states that the results were nearly all within the margin of error of the average of the state-by-state polling data.
Colorado: Clinton had leads in CO for most of the general election campaign although in early September, Trump began posting small leads and in three polls in October and early November, he posted ties.
The final RCP average was Clinton with 43%, Trump with 40.4%.
Clinton received 47.2% with Trump getting 44.4%.
Clinton did better than expected than the average (only two polls out of 30 in the state had her at 48% or above the result she received).
There were five pre-WikiLeaks polls with Clinton receiving anywhere from 48% to 39%.
Maine: ME was only considered a swing state this year because Trump was polling well in the more conservative 2nd Congressional District. Maine awards its 4 Electoral College votes in the following manner: 2 to the statewide winner and then 1 for each CD. Clinton led in all but one of the polls – in mid-September, which was a tie – and finished with an RCP average of 44%, winning by 4.5%. While Trump won a single EC vote due to winning the 2nd CD, Clinton won the state, 47.9% to 45.2%.
In other words, she won with a nearly 4% higher total than she was expected to win by (only one poll, at 48%, out of the 10 tracked in the state had her besting her actual result).
There were no pre-WikiLeaks polls that were tracked by RCP.
Minnesota: MN is never a swing state – it always goes for the Democrats – but it was listed as one after a shocking late-September poll that had the candidates tied. Four other polls showed Clinton leading by 6% to 10%.
RCP didn’t do an average for this state but using a calculator, I came up with 45.8%.
Clinton won the state by about 44,000, 1.5 percent, with 46.9% of the vote compared to Trump’s 45.4%, or about 6-plus point higher than expected.
There were no pre-WikiLeaks polls that were tracked by RCP.
Sidebar: Both Gary Johnson and Jill Stein received about what was expected of them, too, in MN, but, interestingly enough, this is the one state where Evan McMullin, the fringe independent “real conservative” candidate who thought he could take the presidency to the Congress and pull out a win – something that was never going to happen – may have cost Trump votes. He received more than 53,000 votes – or about 9,000 more than the spread between Clinton and Trump. Not unlike the 2000 election though, there’s no guarantee that McMullin’s votes would have gone to Trump; they were “Never Trumpers.” But, with polling so tight in the state, one never knows …
Nevada: This state, where Trump owns a hotel and has employees, was just like FL – seesawing back and forth for most of the campaign – with Johnson polling as high as 10 or 11 points in three polls, even as late as late October.
The RCP average had Trump winning: Trump 45.8, Clinton 45, Johnson 4.
However, the actual results were Clinton 47.9, Trump 45.5, and Johnson 3.3 (with none of the above garnering 2.6%). NV was the only state that predicted Trump would win by the average but he didn’t.
Clinton, again, did better than the average, nearly 3% (only one out of 30 polls, again at 48, had her above what she actually received, again by 0.1%).
There were two pre-WikiLeaks polls: Clinton led the July 7-10 Monmouth poll by 4 points, topping out at 45% or less than her actual Election Day count; Trump led the May 24-25 Gravis poll by 3 points.
New Hampshire: Polling in NH, frankly, was a mess, with an oversampling of Democrats when compared to the actual political makeup of the state in a number of polls. Three polls put together by WMUR and the UNH Survey Center showed Clinton with anywhere from an 11 to 15 point lead; a WMUR poll in late August also showed a 15 point lead. No one believed this polling and most of the other polls, especially in the last week of the campaign, showed the race ranging from Clinton by 1, ties, and Trump by 5, something that was also not believable.
The RCP average had Clinton with 46.8 and Trump with 46.5.
In the end, Clinton won the state in a squeaker – less than 2,800 votes – with 47.6% to Trump’s 47.2%.
Clinton again did better than the average (out of 25 polls, only two had Clinton higher, at 49%, than the vote she received).
There was a single pre-WikiLeaks poll which had the candidates tied with 37%.
Virginia: VA is another previously traditionally Republican state that has now gone purple mostly due to proximity of so many residents to the nation’s capital and the largess that has become Washington, D.C. In every poll – including GOP firms – Clinton led anywhere from 1 to 16 points. Trump pulled resources out of the state in mid-October, according to press reports, and the media began to salivate and continue the phony “Hillary the Inevitable” talking points (a Google search yields nearly 27 million links when searching “Trump campaign pulls out of Virginia”).
Clinton ended with an RCP average of 47.3% while Trump had 44.4%.
On Election Day, she bested that number by close to 3% again, with 49.9% of the vote (only one out of 32 polls had Clinton doing better than her actual results and that was in late October, showing her with 50%). Trump, too, also did better than expected – 45%.
There were three pre-WikiLeaks polls and Clinton led all of them with between 39% and 42%.
Data conclusions on states Clinton won
Even in the states she won, the polling data showed that not only did WikiLeaks have no influence on the outcome, but Clinton performed better than expected.
Other thoughts
This analysis doesn’t delve into the standard things that happen in every election like polling data tightening both at the national level and state levels. It doesn’t take into account that in 2016, the two major party nominees had the highest negatives of any presidential nominees in modern time. There were also very few undecided voters in this election – supporters for both sides were hunkered down, while one side – Trump – was attacked on all fronts, including the press and from within his own party (“Never Trumpers”).
None of this analysis takes into account that in real terms, WikiLeaks was actually under-reported by the national and mainstream press. Yes, bloggers, cable news, and some other outlets were crazed about it and the political industry was obsessed, too, etc. But a very small percentage of the ordinary people of America who bother to vote are connected this nonsense during election cycles; they simply live their lives. There were also no 20 reporters assigned to study the drip, drip, drop, of Wikileaks, like the Washington Post put on investigating Trump (and gathering little from that activity, actually). Any actual in-depth investigation and reporting about the documents released, instead of quick hits, might have truly influenced the election. The press didn’t bother though; many of them had already chosen to be on Clinton’s side. For every “Why are you voting for Trump, he joked about women who let him grab their pussies” question to people admitting they were voting for Trump, there was never a counter question to Sanders’ voters asking, “Clinton and the DNC stole the nomination from Sanders, how can you vote for a woman who rigs an election?”
Most of the Trump-Russian collusion conspiracy coup is based not on facts or evidence but emotion, due mostly to the “Landslide Hillary” media manipulation, that the presidential race in 2016 was in the bag, and the lack of understanding that the daily national tracking poll – and/or the national vote result – has nothing to do with winning the presidency. The U.S. presidential election is a state-by-state battle, to win the most Electoral College votes. There are very few countries in the world that have direct democracies and we aren’t one of them.
The press, bloggers, commentators, and others – on nearly every front – were conditioning the public to believe that Clinton was “Hillary the Inevitable,” that it was going to be a landslide (with the exception of the choke watch, as noted by HeatStreet.com on Nov. 1), etc. How could she lose? Everyone said she would win … the New York Times had her at a 85 percent chance that she would win; FiveThirtyEight.com said it was 71.4 percent; Robert Reich, Bill Clinton’s labor secretary who helped bring on the economic crisis we are in by supporting NAFTA and GATT and then backtracking after the damage had already been done, wrote on Salon.com the day of the election, “Don’t worry”, etc. How did this all happen? It must be … THE RUSSIANS!
There were also independent candidates – a Green and a Libertarian – on many state ballots as well as the delusional write-in candidate McMullin, who assisted in making the election even more unpredictable than it already really was.
Very few people in the press and the politerati predicted that Trump would win. I was one of them (suggesting here and predicting here and here). So was Ann Coulter, on “Real Time with Bill Maher” in June 2015, as she was laughingly mocked at, too, for predicting a Trump win. And there was Michael Moore, just weeks before the election. Even his urging to vote for Clinton couldn’t break from his own reality. These were voices in the blizzard that was the Hillary the Inevitable, the Hillary landslide is coming, etc.
Which brings me back to 2012, again, very quickly. Even though I still think that Trump was the only Republican candidate who could beat Clinton, in actuality, 2016 was always going to be close no matter who the nominees were – because the presidential race in 2012 was actually much closer than most people realize. As I noted in March 2016, Romney barely lost in 2012: 325,000 to 400,000 votes shifted around in five key states and Romney would have won – and he ran a miserable campaign, too, with the press aiding and abetting the incumbent. Obama didn’t win by a landslide in 2012 … he barely won.
All of these reasons – the tricks, the manipulation, the lack of understanding of the process, a Republican nominee who refused to give up, and the volatility of America today – were why Clinton lost the election. Not the Russians.