Slide from Trafalgar Group
Slide from Trafalgar Group

A right wing poll­ster claimed today that Wash­ing­ton’s 2022 con­test for Unit­ed States Sen­ate is “a sur­pris­ing­ly close race” based on a sin­gle find­ing from a sur­vey it con­duct­ed last week that dubi­ous­ly puts Repub­li­can Sen­ate hope­ful Tiffany Smi­ley with­in three points of Demo­c­ra­t­ic incum­bent Pat­ty Murray.

The sur­vey, by Trafal­gar Group, is uncor­rob­o­rat­ed by any oth­er pub­lic opin­ion research and is con­tra­dict­ed by the results of last mon­th’s Top Two, an actu­al elec­tion in which almost two mil­lion Wash­ing­to­ni­ans cast ballots.

Trafal­gar found Mur­ray at 49.2% and Smi­ley at 46.3%, with a mere 4.5% unde­cid­ed. The sam­ple con­sists of 1,087 like­ly gen­er­al elec­tion vot­ers, inter­viewed from August 30th through yes­ter­day, Sep­tem­ber 1st.

Just half a month ago, in the ini­tial round of Wash­ing­ton’s two-part gen­er­al elec­tion, Mur­ray received 52.22% of the vote against Smi­ley and six­teen oth­er chal­lengers. Smi­ley, mean­while, received 33.69% of the vote.

The total vote received by can­di­dates iden­ti­fy­ing them­selves in some form or fash­ion as Repub­li­cans was 41.47%, while the total vote received by Mur­ray and oth­er can­di­dates iden­ti­fy­ing them­selves as Democ­rats was 55.36%. That is typ­i­cal of the split we usu­al­ly see between groups of can­di­dates pro­fess­ing an affil­i­a­tion with the major par­ties in Wash­ing­ton State in a post-2016 Top Two election.

The sec­ond and final round will most like­ly con­sist of an elec­torate that is larg­er, more diverse, and more Demo­c­ra­t­ic, yet Trafal­gar would have us believe that Mur­ray is now crum­bling while Smi­ley is sud­den­ly soar­ing. There is no evi­dence what­so­ev­er to sup­port this con­tention. Trafal­gar­’s sur­vey is a clas­sic outlier.

As I say often here on The Cas­ca­dia Advo­cate, the key to cred­i­ble, accu­rate polling is neu­tral ques­tions asked of a rep­re­sen­ta­tive sample.

Based on what Trafal­gar pub­lished, I have seri­ous doubts that their sam­ple is prop­er­ly rep­re­sen­ta­tive of the like­ly 2022 Wash­ing­ton Novem­ber electorate.

Remem­ber, the gap between Mur­ray and Smi­ley in the Top Two was over eigh­teen points, about in line with what the cred­i­ble pub­lic polling con­duct­ed in the pre­ced­ing month by Elway Research and Sur­veyUSA indi­cat­ed it could be.

This sur­vey puts Smi­ley’s lev­el of sup­port (46.3%) sev­er­al per­cent­age points above and beyond the per­cent­age that every Repub­li­can Sen­ate can­di­date com­bined got in the Top Two. Mur­ray, mean­while, is sup­pos­ed­ly under fifty per­cent despite hav­ing secured more than a major­i­ty of the vote against a field of sev­en­teen oppo­nents as of when the elec­tion was cer­ti­fied on August 16th.

Sor­ry, Trafal­gar, but that dog won’t hunt.

In the gen­er­al elec­tion, there will be no names on the bal­lot besides Smi­ley and Mur­ray. Our own most recent head-to-head poll find­ing, pub­lished at the begin­ning of June, found Mur­ray eleven points ahead of Smi­ley in a hypo­thet­i­cal gen­er­al elec­tion matchup. That was back in the spring, before Mur­ray launched her ads, before the gut­ting of Roe v. Wade, and before the Top Two.

Trafal­gar says on its method­ol­o­gy state­ment page that it con­ducts its sur­veys using live callers, inte­grat­ed voice response, text mes­sages, emails, and “two oth­er pro­pri­etary dig­i­tal meth­ods we don’t share pub­licly,” which is a red flag.

Sub­jec­tive orga­ni­za­tions are per­fect­ly capa­ble of car­ry­ing out objec­tive research, but it appears that Trafal­gar­’s goal here (giv­en their deci­sion to pub­lish these unsup­port­able fig­ures) is to bol­ster Smi­ley’s long­shot can­di­da­cy, rather than mak­ing a con­tri­bu­tion to the body of cred­i­ble pub­lic opin­ion research in this race.

About the author

Andrew Villeneuve is the founder and executive director of the Northwest Progressive Institute, as well as the founder of NPI's sibling, the Northwest Progressive Foundation. He has worked to advance progressive causes for over two decades as a strategist, speaker, author, and organizer. Andrew is also a cybersecurity expert, a veteran facilitator, a delegate to the Washington State Democratic Central Committee, and a member of the Climate Reality Leadership Corps.

Adjacent posts

16 replies on “Poll Watch: Trafalgar tries to bolster Tiffany Smiley’s longshot challenge to Patty Murray”

  1. A 49.2 to 46.3 polling result would mean that King Coun­ty was gross­ly under­sam­pled. Based on recent gen­er­al elec­tion results, the Demo­c­ra­t­ic can­di­date has received between 72 and 75 per­cent in the state’s largest coun­ty, home to 13 of all vot­ers statewide.

    1. That is def­i­nite­ly a pos­si­bil­i­ty, Jer­ry. Trafal­gar did not release a geo­graph­ic break­down of where the respon­dents came from. If that were avail­able we could use it to scru­ti­nize their sam­ple. But it’s sus­pi­cious­ly absent. Also, there’s no 2020 pres­i­den­tial vote break­down in the PDF. Maybe they asked, but if they did, it was exclud­ed from the release.

  2. Why are you so wor­ried about this poll? You try to say it’s uncor­rob­o­rat­ed as well, even though some­one like RCP saw fit to include it in their polling aver­age… (must not be that bad then). Trafal­gar, btw, for some con­text, is the “right wing” poll that showed almost lit­er­al­ly on elec­tion eve that Trump had a shot in Wis­con­sin and Michi­gan, which spurred Trump to go to those states in a last minute swing, lead­ing to his vic­to­ries in both places. Btw, it’s not the first poll to show tight­en­ing. There is anoth­er poll right before this one (look on the RCP polling aggre­gate) that shows that Smi­ley was 6% down rough­ly. This mere­ly rep­re­sents more tightening. 

    The bot­tom line is that you’re only upset about a poll like this if you’re wor­ried. If it’s so obscure, then why talk about it? You talk about it because you’re afraid of the fact that Trafal­gar, while not 100% accu­rate, has shown some lev­el of accu­ra­cy (usu­al­ly with­in a few points) in the past three elec­tions. Be care­ful, Democ­rats, or you’re going to have anoth­er old, worn out ver­sus young charis­mat­ic John James type sit­u­a­tion on your hands.

    1. Mark, I’ve cho­sen to pub­lish your com­ment even though it’s loaded with incor­rect assump­tions, straw­men argu­ments, and con­cern trolling, because we do val­ue dis­sent and wel­come an exchange of ideas here at NPI. 

      You get points for avoid­ing pro­fan­i­ty, although you failed to com­ply with the com­ment­ing guide­line that pro­hibits the use of all caps, so it was nec­es­sary to edit your com­ment before pub­li­ca­tion (the cap­i­tal­ized words were con­vert­ed to boldface).

      Your assump­tion that we are “wor­ried” or “upset” about the poll is incor­rect. As I said, this poll is an out­lier. We have been polling this race our­selves every few months going back to the spring of 2021, and we have con­sis­tent­ly assessed that Tiffany Smi­ley has failed to put Wash­ing­ton in play for the Republicans. 

      With the Top Two elec­tion results now avail­able, we have an even greater degree of con­fi­dence in this assessment. 

      The oth­er poll “right before this one” that you refer to is also a right wing poll; it was con­duct­ed for Jim Jor­dan by the McLaugh­lin Group. Like Tiffany Smi­ley’s inter­nal cam­paign polling from last spring, which also pur­port­ed to show a tight race, it is at odds with the rest of the avail­able body of pub­lic opin­ion research. 

      The clos­est Smi­ley was to Mur­ray in our research was nine points, back in Feb­ru­ary. Our polling indi­cates Mur­ray’s advan­tage has widened since then. 

      The cri­tiques we pub­lish here of the right wing are not lim­it­ed to mat­ters that we are wor­ried about, and nev­er have been. For instance, NPI is well known for pro­vid­ing the likes of Tim Eyman with vig­or­ous, year round oppo­si­tion regard­less of the con­di­tion of his ini­tia­tive fac­to­ry. It’s part of what we do and have been doing for over twen­ty years.

      I nev­er said that Trafal­gar­’s polling was obscure. In fact, Trafal­gar is a well known firm with a large social media pres­ence. Trafal­gar has over 86,000 fol­low­ers on Twit­ter, for exam­ple. A group with that kind of audi­ence is not obscure. 

      It is com­mon for poll aggre­ga­tors like RCP or FiveThir­tyEight to include polls of all sorts: inter­nal cam­paign polls that see a pub­lic release, mass media-com­mis­sioned polls, and polls from sub­jec­tive orga­ni­za­tions with par­ty-aligned poll­sters, such as ours. Just because a poll is includ­ed by an aggre­ga­tor does­n’t make it credible. 

      I men­tioned Tiffany Smi­ley’s own cam­paign polling ear­li­er. That is in the FiveThir­tyEight index, but does­n’t make it any more cred­i­ble than Trafal­gar­’s survey. 

      Every sin­gle time Pat­ty Mur­ray is up, right wing con­cern trolls show up and make com­ments sim­i­lar to your clos­ing line: “Be care­ful, Democ­rats, or you’re going to have anoth­er old, worn out ver­sus young charis­mat­ic John James type sit­u­a­tion on your hands.” It hap­pened in 1998, 2004, 2010, and even 2016. Now it’s hap­pen­ing again. 

      Has it occurred to you that Repub­li­cans are just bad at going up against Pat­ty Mur­ray? Every cycle, it seems, they under­es­ti­mate her, and that works to her advantage. 

      Lin­da Smith was sup­posed to beat her in 1998. Did­n’t hap­pen. Then in 2004, it was gonna be George Nether­cutt, who had knocked out Speak­er Foley in 1994. But nope. Then in 2010, Repub­li­cans recruit­ed Dino Rossi to chal­lenge Mur­ray. Again, she pre­vailed, and in a dif­fi­cult envi­ron­ment for Democ­rats. In 2016, Chris Vance (now an inde­pen­dent) tried to defeat Mur­ray, and got walloped. 

      Here we are in 2022 and Tiffany Smi­ley is run­ning. You imply she’s charis­mat­ic. But she is run­ning a ter­ri­ble cam­paign. She could­n’t even tell The Seat­tle Times’ David Gut­man what she would cut out of the fed­er­al bud­get despite claim­ing to be for rein­ing in spend­ing. The Smi­ley cam­paign con­sists main­ly of boil­er­plate and Biden/Murray bash­ing. We shall see how far that gets Smiley. 

      Our assess­ment, based on the evi­dence: Not far. But any­thing can hap­pen in pol­i­tics. Unlike some, we don’t lack for imagination. 

      For­tu­nate­ly for Democ­rats, Pat­ty Mur­ray nev­er takes a reelec­tion for granted.

  3. First off, the most impor­tant poll will be in Novem­ber. Why do you wor­ry so much about a Sen­a­tor who clings to priv­i­lege and ben­e­fits through five terms, and is run­ning for a sixth? She does not live in Wash­ing­ton State and has not for some time. Mur­ray and oth­er hang­ers-on like her fly in the face of rep­re­sen­ta­tion­al Democ­ra­cy. There are far worse things that can hap­pen to this state and nation than Smi­ley some­how upset­ting Mur­ray. There are two ways to get off the gravy train: To be thrown off or step off. I don’t think Mur­ray is ever going to step off. We lim­it­ed the terms of Pres­i­dents, and until we do it for rep­re­sen­ta­tives and espe­cial­ly Sen­a­tors we dwell in stag­nant polit­i­cal waters.

    1. We aren’t wor­ried about what is going to hap­pen to Pat­ty Mur­ray, Irv­ing. That is an incor­rect pre­sump­tion on your part. 

      You are also laugh­ably wrong about Mur­ray’s res­i­dence. She does live in Wash­ing­ton State. As a Sen­a­tor, she spends plen­ty of time in our nation’s cap­i­tal, but this Wash­ing­ton is where she makes her home. She is usu­al­ly here on the week­ends (even though it’s a long cross coun­try flight) and dur­ing the week when the Sen­ate is not in session. 

      NPI is opposed to term lim­its for mem­bers of Con­gress because impos­ing them could make the “revolv­ing door” phe­nom­e­non even worse. There are more effec­tive ways to tack­le cor­rup­tion and incum­ben­cy, includ­ing com­bat­ing ger­ry­man­der­ing, adopt­ing pub­lic cam­paign financ­ing, and giv­ing cam­paigns access to low cost postage and a min­i­mum amount of free air­time on the pub­lic air­waves to get their mes­sage out.

  4. While I would be shocked if this elec­tion stayed so close, it’s sil­ly of you to dis­miss this poll. Trafal­gar has been one of the most accu­rate polling sites since 2016 when they were one of the few (maybe only?) one to cor­rect­ly call the elec­tion for Trump.

    Also, why do you feel the need to iden­ti­fy them as a “right wing poll­ster”? Oth­er poll­sters you not­ed were only iden­ti­fied by their name rather than as a lib­er­al or inde­pen­dent poll­ster. Sure­ly their polit­i­cal lean must be impor­tant as well?

    Just because you don’t like the results of a par­tic­u­lar poll, or it does­n’t match with your expec­ta­tions, does not mean it must be inac­cu­rate. It seems odd you are going so far out of your way to dis­cred­it this par­tic­u­lar poll.

    1. Bri­an, com­men­tary on polling is stan­dard fare for NPI’s Cas­ca­dia Advo­cate. Our Poll Watch series has been going for years. We pub­lish these kinds of analy­ses all the time. We did not go far out of our way to cre­ate this analysis. 

      I men­tioned in the post that sub­jec­tive orga­ni­za­tions are per­fect­ly capa­ble of con­duct­ing objec­tive research. We won’t dis­miss a poll mere­ly because it hap­pens to have been con­duct­ed by an enti­ty whose lead­er­ship and researchers hold right wing views. But we do dis­miss polling that has either loaded ques­tions or unrep­re­sen­ta­tive sam­ples. Trafal­gar­’s find­ings here are sus­pect because they are not cor­re­lat­ed by any oth­er cred­i­ble data and because Trafal­gar did­n’t pro­vide cru­cial infor­ma­tion need­ed to crit­i­cal­ly eval­u­ate their sur­vey. If you look in the results PDF, you’ll see there’s no geo­graph­ic break­down, there’s no break­down by edu­ca­tion, and there’s no break­down by pres­i­den­tial vote. That’s all very concerning. 

      If Trafal­gar had pub­lished find­ings show­ing Pat­ty Mur­ray forty points ahead, that would also have been sus­pect. Our analy­sis isn’t based on whether we like or dis­like what we see, it’s based on the sound­ness or lack there­of of the ques­tion word­ing, the sam­pling, and the method­ol­o­gy employed. 

      Of the three polling part­ner­ships iden­ti­fied in the post, one has an ide­o­log­i­cal lean (NPI/PPP) and two do not (Crosscut/Elway and SurveyUSA/UW/WSU/KING5/Seattle Times). The one with the ide­o­log­i­cal lean is us. Our ide­ol­o­gy is clear from our name: North­west Pro­gres­sive Insti­tute. That’s not the case with Trafal­gar. So it was impor­tant to men­tion that. Poll aggre­ga­tors like FiveThir­tyEight do the same thing, inci­den­tal­ly: they place sym­bols and labels next to poll­sters inform­ing read­ers if there is an ide­o­log­i­cal lean.

  5. I believe I actu­al­ly respond­ed to that poll. I don’t usu­al­ly respond, but when I heard Trafalager I thought it would be interesting.

    If so, it was def­i­nite­ly a push poll ask­ing ques­tions like ‘Mur­ray has sup­port­ed Biden’s infla­tion­ary poli­cies. How do you feel about her sup­port strong­ly agree to strong­ly dis­agree’.

    Note the sin­gle quotes to indi­cate that wasn’t an actu­al quotation.

    After sev­er­al such ques­tions of the sort the poll asked who I sup­port­ed for Senator.

  6. Queue the panic…Trafalgar is out with anoth­er poll show­ing only a 2 or dif­fer­ence now. Can’t wait to hear why they are wrong again.

    1. Nobody except for Tiffany Smi­ley fans is tak­ing Trafal­gar­’s polling seri­ous­ly, Brian. 

      You came back to the thread, so sure­ly you noticed Jim Love­joy’s com­ment, in which he reports he believes he was part of the sam­ple for the last Trafal­gar sur­vey and detailed that there were a series of loaded ques­tions in there. 

      That helps explain the skew of the results.

      These Trafal­gar polls can’t be trust­ed because: 

      • Their results are con­tra­dict­ed by the body of avail­able evi­dence, includ­ing very fresh research. Elway, which isn’t aligned with any par­ty, just found Pat­ty Mur­ray thir­teen points ahead of Tiffany Smi­ley, which match­es the gap between all the Demo­c­ra­t­ic and Repub­li­can Sen­ate can­di­dates in the August Top Two election.
      • They aren’t releas­ing break­downs of who took their polls by edu­ca­tion, loca­tion, or 2020 pres­i­den­tial vote, which is a red flag.
      • They are not prop­er­ly trans­par­ent about how they’re con­duct­ing their polls… for instance, they claim to use “pro­pri­etary dig­i­tal meth­ods we don’t share pub­licly” for sur­vey­ing peo­ple in addi­tion to emails and texts. Anoth­er red flag.

      Tiffany Smi­ley is not sta­tis­ti­cal­ly tied with Pat­ty Mur­ray and no amount of garbage Trafal­gar data is going to move the nee­dle for her.

  7. I’m not a Smi­ley fan…no idea who she is. I live in Ohio. I just enjoy see­ing lib­er­als freak out. And I’ll keep fol­low­ing the the thread until one of us is forced to admit they were wrong. Talk to you on elec­tion night.

    1. As I men­tioned back in Sep­tem­ber, we nev­er took Trafal­gar­’s garbage polling seri­ous­ly, Brian. 

      But it was impor­tant to call out Trafal­gar for its bad and slop­py work. 

      We now have a first batch of elec­tion results. Pat­ty Mur­ray is up by four­teen points: 

      Pat­ty Murray
      980,199votes (56.92%)

      Tiffany Smi­ley
      738,618 votes (42.89%)

      The elec­tion has been called for Mur­ray and she will be return­ing to the Sen­ate as Wash­ing­ton State’s senior senator.

      Trafal­gar­’s polls in New Hamp­shire, New York, and oth­er states were also way off. They were more inter­est­ed in feed­ing a nar­ra­tive than con­duct­ing legit­i­mate research, which is a shame. 

      We, on the oth­er hand, are com­mit­ted to the sci­en­tif­ic method.

Comments are closed.