There’s some concern in Larry Hogan’s campaign about a New York Times/CBS News/YouGov Battleground Tracker Poll showing Hogan again trails Anthony Brown by double digits, particularly after a Republican pollster showed Hogan trailing by just three points last month. This outfit’s July poll of over 1,400 registered voters showed Brown on top by 13 points in July, and is computed in the Real Clear Politics average.
The Republican’s campaign contends the poll is “so flawed and so misleading that Politico hammered the New York Times for lending their name to this internet survey.” YouGov’s methods are different than most pollsters, as they conduct their surveys among an audience which opts into the poll via the internet, then weighs the results among demographics. The overall survey even solicits new customers, requesting people to “Join YouGov today to take part in surveys like these and earn money…”
Maryland’s race is also interesting because of the relative lack of responses compared to other states. Out of 35 states surveyed, Maryland only beats 12 states in terms of participation. Most of the states Maryland beats are fairly rural and sparsely populated: Alabama, Hawaii, Idaho, Kansas, Nebraska, Oklahoma, Rhode Island, South Carolina, South Dakota, Tennessee, Vermont, and Wyoming. It’s not a tremendously representative sample.
Nor is it necessarily reflective of the Maryland electorate. The unweighted sample actually has independents well over their voting strength at 29%, with Democrats comprising 45% and Republicans only 25%. (It’s actually close on the GOP.) But weighting the sample as YouGov does places the Democrats at 52%, independents at 26%, and Republicans at only 22%. In reality, according to the latest voter registration figures, Democrats have 55% share, Republicans 25.7%, and independents just 19.3%. So both major parties are undersampled by about 3% apiece.
Polling is all about turnout. While the YouGov survey claims these are “likely voters,” in reality those not affiliated with a party are the least likely to turn out for a gubernatorial election. Yet when I reset their polling numbers to a very likely turnout model (that of the 2010 election, which was a muted TEA Party wave election in the state) and distribute the “not sure” voters in the same proportion as those who have decided, I come out with this possible result:
- Brown 57.5%, Hogan 40.2%, other 2.3%
I think the reason this turns out the way it does is that the YouGov sample has Brown winning Democrats at roughly the same rate Hogan wins GOP voters. In a lot of ways the YouGov poll is almost a worst-case scenario for Hogan, who needs to both boost turnout for his side to levels last seen in 2002, when almost 68% of Republicans and over 45% of independents came out to vote – in 2010 those numbers were about five points lower – and get far more than the 6% of Democrats the YouGov poll has voting for him. If Anthony Brown can convince Democrat voters to stay loyal to the nominee, the game is over, and that’s why Brown’s going negative.
In fact, Hogan’s campaign added that:
If the MD Democratic Party – with their two-to-one registration advantage over Republicans – honestly thought Brown was ahead, they wouldn’t need O’Malley’s Democratic Governors Association to spend $750,000 in special interest money on attack ads to bail out his campaign.
So I think the reality is somewhere between the 14 points this poll has Brown leading by and the 3 points Hogan claims he is behind. It just proves there’s a lot of work to do in explaining the real record of Anthony Brown and the damage his policies would do to Maryland if he’s elected.