A New Jersey subsample of a national poll conducted by Ipsos found less support for marijuana legalization than other recent polls on the issue, but the results don’t come without baggage.
About 38% of New Jersey registered voters polled said they supported legalization. Another 26% backed decriminalization while 25% wanted no change in marijuana laws, and 10% backed prohibitions on all types of marijuana sales.
The problem there, said Ashley Koning, director of Rutgers University’s Eagleton Poll, is that those figures are all drawn from the same question.
“You never want to have something that’s double-barreled, where you’re asking people two, three, four questions within the same question stem,” Koning said. “It’s confusing to the respondent and then you don’t really know what the response implies if you’re asking them four different questions but you’re asking them to only give you one answer back.”
An Eagleton poll conducted at the end of October found significantly more support for marijuana legalization than the Ipsos poll.
In the October poll, 33% said they strongly supported legalization, and 25% said they somewhat supported the same. Another 25% were strongly opposed while 12% were somewhat opposed.
“This poll continues to prove that, given the option between all possible options of drug policy, people want to see something other than full legal sales of marijuana,” said Kevin Sabet, president of Smart Approaches to Marijuana, which is pushing the Ipsos poll. “Other polls on this subject offer a false dichotomy that the only options here are either full legalization or full prohibition and that could not be further from the truth.”
Sabet served as a National Drug Control Policy advisor for Presidents George W. Bush, Barack Obama and Bill Clinton.
SAM opposes marijuana legalization but backs civil over criminal penalties for marijuana use while supporting criminal charges for production and sale of cannabis.
New Jersey lawmakers have been pushing a marijuana legalization bill, but that effort recently stalled after it became unclear whether they would have the votes to pass the legalization package, which includes bills on medical marijuana expansion and expungement for marijuana offenses.
The Ipsos poll’s New Jersey figures are drawn from a larger, nationally-representative poll of about 1,000 respondents, but because of that, the sample size for the state numbers is unusually low. The New Jersey sample had 304 respondents, of whom 266 were registered voters.
The latter figures fall well short of an ideal sample size, Koning said.
“A sample size of 300 doesn’t allow you to look at those breakdowns as much, and any number less than 100, you’re having a 10-point or more margin of error on those kinds of results. So if you think about it, these results could be what they’re saying — this estimate — they could be 10 points higher, they could be 10 points lower. The confidence interval and the margin of error just blow up the smaller your sample size is.”
A spokesman for SAM said the subsample had a 5.6% margin of error.
This story has been updated with a margin of error for the subsample.