Fact-check­ers in heated de­bate on nonci­t­i­zen vot­ing

The Washington Times Weekly - - Politics - BY ROWAN SCAR­BOR­OUGH

A right-lean­ing fact-checker is fight­ing crit­ics on the left who say its con­clu­sion that a lot of nonci­t­i­zens vote il­le­gally is bunk.

The on­line bat­tle of de­bunk­ing and re­but­tal is play­ing out as a much larger war has erupted be­tween Pres­i­dent Trump’s com­mis­sion on elec­tion in­tegrity and Demo­cratic state lead­ers. They are re­fus­ing to pro­vide the panel with pub­lic voter reg­is­tra­tion data. Left-wing groups are su­ing to stop the com­mis­sion’s work, which could set­tle the nonci­t­i­zen de­bate by col­lect­ing enough data.

In the fact-check stand­off, there is Just Facts, a small New Jersey firm of con­ser­va­tive and lib­er­tar­ian schol­ars who pro­mote what they say is solid in­de­pen­dent re­search.

Just Facts Pres­i­dent James D. Agresti is­sued a block­buster re­port in June. Us­ing pre­vi­ous re­search, polling data and Cen­sus Bu­reau fig­ures, his team con­cluded that as few as 594,000 nonci­t­i­zens or as many as 5.7 mil­lion voted in the 2008 pres­i­den­tial elec­tion. If ac­cu­rate on the high side, it would vin­di­cate Mr. Trump’s con­tention that a lot of il­le­gal bal­lots were cast in his race for the White House last year with Demo­crat Hil­lary Clin­ton.

Chal­leng­ing Mr. Agresti are fact-check­ers Poli­tiFact and Snopes.com, which con­ser­va­tives gen­er­ally view as lib­eral; lib­eral col­lege pro­fes­sors; and left-lean­ing news sites such as Huf­fPost.

When Mr. Agresti read their broad­sides, he is­sued a re­but­tal that es­sen­tially said they do not un­der­stand sta­tis­tics. Sci­en­tific ran­dom sam­ples, though small, can be used to es­ti­mate be­hav­iors of larger pop­u­la­tion sec­tors as all polls do, he said.

Poli­tiFact wrote: “The num­ber comes from a con­clu­sion by Just Facts, a con­ser­va­tive/lib­er­tar­ian think tank. Just Facts’ num­bers came from a study by Old Do­min­ion Univer­sity re­searchers. That study was based on a sur­vey which showed that 38 peo­ple out 32,800 claimed to be nonci­t­i­zens who had voted. Just Facts used data from the study and cen­sus es­ti­mates on the nonci­t­i­zen pop­u­la­tion to come up with a na­tional fig­ure of nonci­t­i­zen vot­ers.

“But other re­searchers and po­lit­i­cal sci­en­tists have said the small num­ber is not a re­li­able source of data on nonci­t­i­zen vot­ers na­tion­wide. We rate this claim False.”

Mr. Agresti wrote in his re­but­tal that his crit­ics’ ar­ti­cles “are lit­tered with sopho­moric in­con­sis­ten­cies, ir­ra­tional­i­ties and out­right false­hoods.”

The ref­er­enced sur­vey is the Co­op­er­a­tive Con­gres­sional Elec­tion Study spear­headed by Har­vard Univer­sity. Its study, with polling by YouGov.com, is the data­base used by Old Do­min­ion Univer­sity schol­ars who is­sued the first ex­ten­sive study of its kind on how many nonci­t­i­zens vote il­le­gally.

The left, whose re­searchers and politi­cians are con­fi­dent that few aliens vote, im­me­di­ately scorned the prod­uct from Old Do­min­ion.

‘Math­e­mat­i­cally il­lit­er­ate no­tions’

The lib­eral press de­clared the ODU study de­bunked even though the authors is­sued a pa­per de­fend­ing their work. A group of lib­eral pro­fes­sors cir­cu­lated a let­ter call­ing on col­leagues to black­list the study.

The de­bate is com­plex be­cause it is in­ter­twined in polling data, cen­sus num­bers, ac­tual recorded voter be­hav­ior and as­sump­tions about how peo­ple an­swer sur­veys.

But there is a cen­tral fact: The Har­vard study polled 32,800 peo­ple. Of those, 339 were self-de­scribed nonci­t­i­zens, and of those, 38 said they def­i­nitely voted.

The Har­vard re­searchers said the sam­ple is too small to draw con­clu­sions. They de­con­structed the 38 fig­ure and elim­i­nated all of them based on a com­par­i­son with voter data­bases and other fac­tors.

Poli­tiFact, Snopes and lib­er­als as­serted that the Har­vard re­searchers de­bunked ODU and Just Facts.

Poli­tiFact points to Brian Schaffner, a po­lit­i­cal sci­ence pro­fes­sor at the Univer­sity of Mas­sachusetts at Amherst and an au­thor of the Har­vard study. It quoted his Politico col­umn: “Sur­vey re­spon­dents oc­ca­sion­ally se­lect the wrong re­sponse by ac­ci­dent — per­haps be­cause they are rush­ing through and not read­ing the ques­tions care­fully, be­cause they do not fully un­der­stand the ter­mi­nol­ogy be­ing used, or be­cause they sim­ply click on the wrong box on the page.”

Mr. Agresti said the Old Do­min­ion schol­ars took the is­sue of sam­pling into con­sid­er­a­tion by “weight­ing” the data to match Cen­sus Bu­reau num­bers.

In ad­di­tion, he said, 9 per­cent of self­de­clared nonci­t­i­zens who said they were not reg­is­tered to vote were in fact on voter rolls, ac­cord­ing to data­bases.

Of those, 8 per­cent an­swered, “I def­i­nitely voted” in 2008, and of those, 82 per­cent said they voted for Barack Obama.

“Given the poor math skills of many adults, the ar­gu­ments of Snopes and Poli­tiFact may sound con­vinc­ing to some peo­ple, but these so-called fact-check­ers are us­ing math­e­mat­i­cally il­lit­er­ate no­tions in­stead of con­crete, quan­tifi­able facts,” Mr. Agresti wrote.

As back­ground, Mr. Agresti points to a sci­en­tific sur­vey by the mar­ket­ing con­sul­tancy McLaugh­lin & As­so­ci­ates. In a poll of His­pan­ics for the Cal­i­for­nia wine in­dus­try, 13 per­cent of nonci­t­i­zens said they were reg­is­tered to vote. Matched against the coun­try’s nonci­t­i­zen pop­u­la­tion, that means 800,000 to 2.2 mil­lion aliens were reg­is­tered to vote, Just Facts con­cluded.

About 20 mil­lion nonci­t­i­zens, in­clud­ing 11.8 mil­lion His­panic adults, are liv­ing in the U.S., the Cen­sus Bu­reau said in 2013.

The com­mis­sion’s mis­sion

Mr. Agresti said it is more likely that nonci­t­i­zens said they were cit­i­zens in the Har­vard sur­vey, not the other way around, be­cause tech­ni­cally they oth­er­wise would be ad­mit­ting to a crime.

“More­over, even if Just Facts dis­re­garded all nonci­t­i­zens who said they voted and lim­ited its anal­y­sis only to votes con­firmed by vot­ing records, the data would show that 590,000 to 3.9 mil­lion nonci­t­i­zens voted in the 2008 elec­tion,” he said. “In ac­cord with sci­en­tific re­search meth­ods, Just Facts fully ac­counts for the sam­pling er­ror in its study, and this is why the re­sults span a wide range of 594,000 to 5.7 mil­lion il­le­gal vot­ers.”

Poli­tiFact said it con­ferred with three elec­tions an­a­lysts out­side the Har­vard study who agreed with the de­bunkers.

Mr. Agresti said that while Poli­tifact cast him as an ide­o­logue con­ser­va­tive and crit­ics as po­lit­i­cally pure, it failed to dis­close that two of the quoted an­a­lysts con­trib­uted money to the Obama cam­paign. The third wrote an op-ed urg­ing mil­len­ni­als not to vote for a third party be­cause it would help Don­ald Trump.

One way to set­tle the nonci­t­i­zen de­bate could be the Pres­i­den­tial Ad­vi­sory Com­mis­sion on Elec­tion In­tegrity. The co-chair­men, Vice Pres­i­dent Mike Pence and Kansas Sec­re­tary of State Kris Kobach, want the states to pro­vide what is nor­mally pub­lic voter reg­is­tra­tion data.

The com­mis­sion would com­pare those lists with other data, such as So­cial Se­cu­rity rolls and Home­land Se­cu­rity De­part­ment files on nonci­t­i­zens who hold green cards.

In 2013, the So­cial Se­cu­rity Ad­min­is­tra­tion found that 700,000 il­le­gal im­mi­grants were us­ing So­cial Se­cu­rity num­bers with fraud­u­lent birth cer­tifi­cates and 1.8 mil­lion had cards that did not match their names.

Demo­crat-led states are stonewalling the com­mis­sion, and a leader of this re­sis­tance is Vir­ginia Gov. Terry McAuliffe.

He has ve­toed a string of bills de­signed to weed out il­le­gal vot­ers. His ap­pointees also broached the idea of mak­ing op­tional the cit­i­zen­ship ques­tion on voter reg­is­tra­tion forms but aban­doned the idea amid strong op­po­si­tion from Repub­li­cans.

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.