Independent New Hampshire Journalism · Northfield, NH
Robert Greene’s thirty-third law says everyone has a weak point, and whoever finds it holds a lever. Two New Hampshire debates and one court case show the finding is easy. The record is short on finders who came out ahead.
On the night of Feb. 6, 2016, at Saint Anselm College in Manchester, seven Republicans debated three days before the New Hampshire primary. Marco Rubio was coming off a strong finish in Iowa. A week earlier, NBC News reported, he had seemed better positioned than ever to pull the party’s mainstream behind him. Chris Christie, the governor of New Jersey, went straight at him. Rubio, Christie said, did not have the experience to be president. Rubio answered with a line about Barack Obama. Then he gave it again.
“There it is,” Christie said. “The memorized 25-second speech. There it is, everybody.”
Robert Greene’s thirty-third law says to discover each man’s thumbscrew: everyone has a weak point, and whoever finds it holds a lever. New Hampshire’s primary is where candidates go looking. The record has a Republican who found a rival’s weak point on live television and a Democrat who pressed one eight years earlier on the same stage. It also has the industrial version, the phone call that tests an opponent’s weaknesses while posing as a poll, and a state law written to unmask it. In each case the weak point was real. What the record does not show is the person who found it collecting the reward.
The stage
Greene’s law, as summaries of the book put it, treats weakness as universal. Everyone has a gap in the armor. Find it and you have a hold on the person.
The New Hampshire primary is built for the search. Candidates spend months in small rooms and then meet on one stage days before the vote, with every campaign having studied every other. A debate days before the primary has been held at Saint Anselm more than once, and twice in eight years it turned on a single exchange about a single flaw.
Manchester, 2016
Rubio’s flaw, as Christie framed it, was that he was scripted. The senator’s line that night was “Let’s dispel with this fiction that Barack Obama doesn’t know what he’s doing.” CBS News reported that he repeated it at least twice more within the next five minutes. Christie told the audience the line was a memorized speech handed to Rubio by his advisers, and said that was how Washington worked. Rubio hit back that New Jersey’s credit rating had been downgraded nine times. The Associated Press account ran under a headline that had Rubio stuck in a loop and Christie on the attack.
Christie had a second line of attack ready. He went after Rubio’s missed votes in the Senate, including one on sanctions against Hezbollah, and called the absences truancy. But it was the repetition that stuck, because Rubio supplied the evidence himself, in real time, while the charge was still in the air.
The screw had been found in public, and voters said they noticed. In exit polling, almost 70 percent of Republican primary voters said the debate played an important role in their decision, CBS News reported. Nearly 90 percent called it at least a factor. The voters who rated it important were less likely to support Rubio than John Kasich.
On primary night Rubio did not argue with the verdict. “I did not do well on Saturday night,” he told his supporters, and promised it would not happen again.

Then look at who gained. With about 73 percent of precincts reporting, The Washington Times had Donald Trump at 35 percent and Kasich at 16. Rubio was fifth at 10. Christie was sixth, a little under 8. The next day Christie suspended his campaign. He said he left the race “without an ounce of regret,” CBS News reported.
Christie found the weak point, named it and proved it in front of the electorate. The benefit went to candidates who had not said a word in the exchange.
The same stage, 2008
Eight years earlier the Democrats stood in the same hall. It was Jan. 5, 2008, three days before the primary. Hillary Clinton had just finished third in Iowa, and a CNN, WMUR and University of New Hampshire poll had Obama ahead of her in New Hampshire, 39 percent to 29, The Dartmouth reported.
The moderator, Charles Gibson of ABC, asked Clinton about her weak point directly: her poor likability ratings with voters. She said that hurt her feelings and that she would try to go on. She called Obama very likable and said she did not think she was that bad.
Obama did not let it pass. “You’re likeable enough,” he said.
Three days later Clinton won the primary, 39 percent to 37, defying polls that had predicted a decisive Obama victory, The Dartmouth reported. Her strength among women and older voters helped carry her. In her victory speech she said she had listened to New Hampshire that week, “and in the process I found my own voice.”

No source read for this piece proves the remark moved votes, and the piece does not claim it did. The sequence is what the record gives: the weak point was named on stage, the front-runner leaned on it, and the candidate who was leaned on won.
The call that sounds like a poll
A debate is the retail way to find a weak point. Campaigns also do it wholesale, by telephone. A caller who sounds like a pollster asks a voter a series of questions about a candidate, and the questions carry the attack. The campaign learns which weakness lands, and the voter hears all of them.
New Hampshire has had a law about that since 1998. It does not ban the practice. It makes the caller say who is calling.
The definition is in RSA 664:2. Push-polling means contacting voters for or against a candidate, asking questions about an opponent that “state, imply, or convey information” about that person’s character, status, stance or record, and doing it in a way a voter would likely take for an independent survey. The current text reaches text messages, the internet and social media as well as the phone.
The third part of that definition is the one that matters. A campaign that calls under its own name and attacks its opponent is running an ad. The law is aimed at the call that borrows the neutrality of a pollster, so the person on the line weighs the claims as information and not as a pitch.
The law met its limit in 2014. In September 2010, the attorney general’s office received information about polling calls to New Hampshire residents that were described as carrying negative content about Ann McLane Kuster, a candidate for Congress. The office investigated and concluded that the Bass Victory Committee, the campaign committee of former U.S. Rep. Charles F. Bass, had engaged in push-polling without the required disclosure. It went to Superior Court for civil penalties.
The version of the law then in force asked for less than the current one. A caller had to say the call was being made for or against a particular candidate, name the candidate and give a telephone number. The attorney general alleged the committee’s calls came without that disclosure.
The committee never had to answer on the facts. It argued that federal campaign finance law overrides state law on elections to federal office, and the trial court agreed and dismissed the case. Its reasoning was short: a campaign has to spend money to make the calls, federal law governs what a federal campaign must disclose about its spending, and so the state could not add a disclosure of its own. On Oct. 15, 2014, the New Hampshire Supreme Court affirmed. The disclosure rule, the court held, is a requirement placed on a campaign expenditure, and federal law preempts it “as applied to federal candidates and political committees.”
No court decided whether the calls were push polls. The ruling took the statute off the table for federal candidates and their committees. It did not touch state and local races.
What the three cases share
In all three, finding the weak point was the easy part. Rubio’s reliance on a script, Clinton’s likability numbers and whatever the 2010 callers asked about were not secrets. Rival campaigns and pollsters already had them.
The hard part was the turn. Christie turned it in person, with his own name on it, and both men finished out of the top four. Obama turned it with three words and lost the primary three days later. The phone version exists because it solves that problem: the caller has no name, so the cost of the attack lands on no one. That is the piece New Hampshire’s statute goes after. It does not protect the candidate’s weak point. It takes away the caller’s cover.
That is also why the 2014 ruling matters more than its narrow holding suggests. In a race for governor, state senator or selectman, a New Hampshire voter is entitled to hear who is behind the questions before the first one is asked. In a race for Congress or president, the same voter is not, at least not under this statute.
A fair objection
The first objection is that Christie’s attack worked. Rubio fell from contender to fifth, and that was the point of it. That is fair as far as it goes. Greene’s law promises an advantage to the person who finds the screw, and Christie was out of the race four days after he turned it.
The second objection is that testing an opponent’s weaknesses is ordinary campaign research, and that New Hampshire’s law sweeps too broadly. The record supports part of that. The definition was amended in 2014, and the current text sets thresholds, such as more than 2,000 calls of under two minutes each in the largest races. The Federal Election Commission concluded in a 2012 advisory opinion that the state rule could not be applied to federal candidates, the Supreme Court noted, and the court agreed.
The third is that one debate does not decide a primary. Also fair. The exit poll says the 2016 debate mattered to most voters. Nothing read here measures the 2008 remark.
The verdict
Greene’s thirty-third law is half right in New Hampshire. Weak points are there to be found, and the primary all but hands them out. The law’s promise is the other half, that the finder gains, and on this record the finder mostly did not. Christie finished sixth. Obama lost a primary he was favored to win. The committee accused of the anonymous version won in court on federal preemption, with the facts never tried.
The people who gained were standing nearby. Kasich took second place in 2016. Clinton took the 2008 primary after her weak point was read aloud on stage.
Part Thirty-Two showed sellers who prospered by offering a dream. Part Thirty-One showed a candidate who won by writing the question. Part Thirty-Three shows the limit of going after the other person directly: in a small state, the voters watch the hand on the screw as closely as the person under it.
Next: Law 34, on acting like a king to be treated like one, tested against the New Hampshire record.
Fact check
| # | Claim | Status | Source |
|---|---|---|---|
| 1 | Feb. 6, 2016, debate at Saint Anselm College in Manchester, sponsored by ABC News, three days before the primary; seven candidates; Rubio coming off a strong Iowa finish; Christie on Rubio’s experience and “There it is. The memorized 25-second speech. There it is, everybody.”; Christie called Rubio’s missed Senate votes truancy; headline “Rubio stuck in a loop, Christie on the attack” | ATTRIBUTED | Steve Peoples and Julie Bykowicz, The Associated Press, Feb. 6, 2016 (as published by PBS NewsHour) |
| 2 | Rubio’s line, “Let’s dispel with this fiction that Barack Obama doesn’t know what he’s doing”; repeated at least twice more within five minutes; Christie said the line came from Rubio’s advisers and that this is what Washington does; Rubio’s reply on New Jersey’s nine credit downgrades; Christie on the missed Hezbollah sanctions vote | ATTRIBUTED | Rebecca Kaplan and Reena Flores, CBS News, Feb. 7, 2016 |
| 3 | A week earlier Rubio seemed better positioned than ever to consolidate mainstream Republican support | ATTRIBUTED | NBC News, primary-night report, Feb. 9, 2016 (byline and date not shown in the page text) |
| 4 | Exit polling: almost 70 percent said the debate played an important role and almost 90 percent called it a factor; those voters were less likely to support Rubio than Kasich; Rubio’s “I did not do well on Saturday night”; Christie returning to New Jersey | ATTRIBUTED | Rebecca Kaplan, CBS News, Feb. 9, 2016 |
| 5 | With about 73 percent of precincts reporting: Trump 35 percent, Kasich 16, Cruz 12, Bush 11, Rubio 10, Christie slightly less than 8 | ATTRIBUTED | The Washington Times, Feb. 9, 2016 (partial returns) |
| 6 | Christie suspended his campaign Feb. 10, 2016, after a sixth-place finish with single-digit support; “without an ounce of regret” | ATTRIBUTED | Reena Flores, CBS News, Feb. 10, 2016 |
| 7 | Jan. 5, 2008, debate at Saint Anselm, sponsored by ABC, WMUR and Facebook, moderated by Charles Gibson; the likability question; Clinton said that hurt her feelings and remarked on Obama; Obama’s “You’re likeable enough”; Clinton third in Iowa; CNN/WMUR/UNH poll, Obama 39, Clinton 29 | ATTRIBUTED | William Schpero, The Dartmouth, Jan. 7, 2008 (UWIRE, as published by CBS News) |
| 8 | Clinton 39 percent, Obama 37; polls had predicted a decisive Obama win; strength among women and older voters; “and in the process I found my own voice” | ATTRIBUTED | Allie Lowe, The Dartmouth, Jan. 9, 2008 (UWIRE, as published by CBS News) |
| 9 | RSA 664:16-a text: “paid political advertisement,” identification of caller and payer, telephone number, candidate named; source notes from 1998, 12:3 through 2021. RSA 664:2, XVII: definition of push-polling, including “state, imply, or convey information,” other media, and the batch thresholds | VERIFIED | gc.nh.gov, RSA 664:16-a and RSA 664:2 |
| 10 | September 2010 information to the attorney general about calls with negative content about Kuster; the office’s conclusion about the Bass Victory Committee; petition for civil penalties; the disclosure the statute then required; the trial court’s reasoning that push-polling is a campaign expenditure; dismissal on preemption; the FEC’s 2012 advisory opinion; decision of Oct. 15, 2014, affirming; preemption “as applied to federal candidates and political committees”; statute amended 2014 | VERIFIED | New Hampshire Attorney General v. Bass Victory Committee, 166 N.H. 796 (2014), opinion text read on CourtListener |
| 11 | Featured photograph: statue of John Stark on City Hall Plaza in Manchester, Carol M. Highsmith, LC-DIG-highsm-47333, no known restrictions | VERIFIED | Library of Congress catalog record 2017883964 |
| 12 | Robert Greene’s Law 33 is “Discover Each Man’s Thumbscrew,” from The 48 Laws of Power, Viking, 1998 | VERIFIED | Shortform summary of the book, read this session |
| 13 | GSR Power Index scores for Laws 1 through 33 | EDITORIAL | Granite State Report’s own rating, not a measurement. Method stated in the chart |
Your voice
Poll. When one candidate exposes another’s weakness in a debate, who gains?
The attacker · The target, from sympathy · Someone else in the race · No one
Survey: the call that wasn’t a poll. Five questions: your town; whether you have received a political call that sounded like a poll and pushed negative claims; about what year; whether the caller said who paid for it; and what they asked. Granite State Report asks before using your name or story.
Discuss. Is going after a rival’s weak point fair campaigning, or does it tell you more about the attacker? And should a caller testing attack lines have to say who is paying?
Verified New Hampshire residents vote and answer at granitestatereport.com. Or write to granitestatereport@gmail.com with your town in the subject line.
Granite State Report wants to hear from voters who received one, and from campaign staff and pollsters who have written or refused to write them. granitestatereport@gmail.com, 603-931-9264. Confidentiality on request.
Editor’s note: Commentary and analysis, not legal advice. The court opinion and the two statute sections were read directly. The attorney general’s conclusions about the 2010 calls were allegations; the case was dismissed on federal preemption and no court ruled on them. Accounts of the 2016 and 2008 debates come from contemporary reporting and are attributed. The 2016 vote shares are partial returns as reported on primary night, and the 2008 figures are as reported by The Dartmouth; neither is the official canvass. The piece draws no causal link between the 2008 debate remark and the result. The GSR Power Index is editorial, not data.

