2008 New Hampshire primary

Jump to: navigation, search

The 2008 New Hampshire primary was expected to be swept by Barack Obama after his major win in Iowa. Pre-election polls predicted a large Obama victory, as did the exit polls. But on election day, Hillary Clinton won a shocking upset instead. Her victory was ascribed to political shifts that the polls missed, but evidence points to election fraud being the true cause.

Pre-election polls

Clinton led most of the 2007 New Hampshire polls, but Obama began closing the gap in early 2008. He was bolstered by his victory in the Iowa caucus, which shot him up to a significant lead. In the week before the election, Obama led Clinton by anywhere from 7% to 13%. The final RCP average was an 8.3% Obama lead.[1] Yet Clinton won by 2.6%, an 11% reversal of the pre-election polling.

In fact, it wasn't just public polls showing Obama ahead. The internal polls of both Obama and Clinton's campaigns projected Obama winning.[2]

Exit polls

The exit polling for New Hampshire also showed Obama significantly ahead[2], leading Clinton by 5%.[3] Clinton's win, then, was nearly an 8% reversal of the exit polls.


Essential victory

New Hampshire was clearly a must-win state for Clinton. She had just been humiliated by two big losses to Obama, the charismatic newcomer who unseated her from the lead. If Clinton lost New Hampshire, Obama's momentum would likely become unstoppable and sink her candidacy. If Clinton won, she'd remain viable in the primaries. For anyone who cared about keeping Clinton in the running, it was crucial to ensure she won New Hampshire, by any means necessary.

Shocking upset

All of the polls going into election day had Obama winning easily. This included not just pre-election polls, meant to predict the electorate, but exit polls, taken of actual voters as they left the polling place. All of these different polls agreeing on a large Obama victory made Clinton pulling off a win all the more shocking. Clinton's surprising upset reversed the polls by anywhere from 8% to 16%, a massive shift.

To explain the pre-election polls missing, pundits pointed to political factors. They brought up Clinton's passionate defense of her candidacy, her display of emotion, and debate gaffes by Obama leading up to the New Hampshire primary. But even if these reasons justify the large shift from the pre-election polls, they don't explain why the exit polls (of voters who actually turned out) gave Obama a similar win.

Standard arguments against exit polls, like overweighting certain demographics, are also unlikely. Exit pollsters know how to account for demographics, and similar arguments for past elections have been consistently debunked. But with Obama's race, the Bradley effect might have also been in play. It posits that people will lie to pollsters about supporting a black candidate so as to not seem racist, despite having chosen someone else. While the Bradley effect could account for the exit poll discrepancy, there are two reasons to doubt it: the massiveness of the shift (8%) and the fact that exit poll questionnaires are confidential written surveys, removing an element of societal pressure.

While political and social factors could explain the polling discrepancies, the New Hampshire results still stand out as strange. Virtually every poll, including one of real voters who turned out, was in agreement, yet the electoral outcome was different. And the benign explanations for the poll misses are fairly weak.

Hand vs. machine counts

New Hampshire's towns were split between using optical scanners or hand counting to count paper ballots. Hand-counted paper ballots, whose count can be publicly observed, are considered the gold-standard of election accuracy. Unlike voting machines, which conduct counts invisibly, hand counts are out in the open for everyone to see, leaving little room for manipulation.

A useful metric, then, for checking unverifiable optical scanner counts is to compare against the hand counts. Obama won hand-counted towns by 4.1%, but lost machine-counted towns to Clinton by 4.4%.[4] The fact that Clinton did 8.5% better when votes were counted on unobservable machines prone to manipulation, and how the hand-count and machine-count results were essentially mirror images, is rather telling. It hints at voting machine manipulation in Clinton's favor.

Of course, there could have simply been differences between hand-counted and machine-counted towns that explain the disparity. But Bev Harris, an election integrity expert, has noted that the hand-counted areas are fairly representative of the state, and statistical analysis debunks demographic reasons for the disparity. One factor that can be ruled out immediately is population: among both small and medium towns (there weren't enough large hand-counted towns), Clinton does about 8% better with machine counting.[4] A further analysis finds that even when controlling for numerous town demographics - population, education, income, age, geography, competitiveness - Clinton does much better when machine counts are employed.[5]

The unexplained link between Clinton's performance and the use of unaccountable and vulnerable machines is hard to justify innocently. It strongly indicates that the voting machines were rigged in her favor.

Untrustworthy contractor

Main article: LHS Associates

LHS Associates, responsible for programming all the voting machines in New Hampshire, is quite the opposite of a reputable contractor. Its employees (some of whom are criminals) show an extreme disdain for election integrity: deriding concerns over voting machine security, violating election law, and even threatening to rig elections. The company is very secretive, programming the machines with no public oversight, and its documents aren't subject to public records laws. Anyone at LHS Associates could rig an election without being caught, and the criminal elements within the company make that a realistic possibility.


A recount, called for by Dennis Kucinich, didn't change the results. However, Bev Harris and other observers documented so many chain-of-custody and integrity violations that the recount couldn't possibly be considered honest[6][7]:

  • Local election offices, where the paper ballots were stored, had suspicious nighttime activity the weekend before the recount
  • Ballot boxes, with flimsy seals, arrived already open. Some of them were deliberately left overnight in an unsecured room. In one case, the Manchester Ward 5 ballots were in the Ward 6 box and vice-versa, indicating someone messed with the ballots and returned them to the wrong box.
  • An election official was pulling a results sheet out of his pocket and whispering to the counting teams
  • 20-or-so consecutive ballots had identical marks, as if they were produced by a copy machine
  • While the daytime recount was going on, a group of election officials had the ballots out at night. There were several ballot boxes open on the table, and five cans of acetone (ink solvent) underneath it.

All of these violations point towards ballot and procedural manipulation to match a predetermined result. And it begs asking why the recount was fraudulent if the initial vote counts were correct.


Clinton's win ensured that the 2008 Democratic primaries would be a protracted battle, averting Obama's expected clinching of the contest. Numerous groups benefited from this: Clinton's own campaign, a media with a months-long dramatic race on their hands, and Republicans happy to let Democrats divide themselves in a contentious primary fight. The signs of election irregularities were ignored, and a corrupt recount only reaffirmed that everything was fine.


  1. Obama substantially led pre-election polls
  2. 2.0 2.1 Exit polls had Obama winning
  3. Obama leads exit poll by 5%
  4. 4.0 4.1 Hand and machine count comparison
  5. Machine use tied to Hillary's performance
  6. Bev mentions terrible recount chain of custody
  7. Recount integrity violations

External links

Statistical analyses

Ballot integrity and recount