cannabineer
Ursus marijanus
I dislike polls being presented as news. Most readers don’t even know how they’re being manipulated.Another example of polling distortions introduced by habit & method.
one of the most UNexamined sources of distortion is telephone culture.
we have two distinct such: cell culture & wire culture
Wire culture uses POTS (plain old telephone service) which is wired into the structure of home and office. The people who still maintain POTS lines have a culture of ANSWERING THE PHONE. EVERY time & quickly, if possible. As a result, they answer the phone prepared to interact & respond. They tend to be on average older, more rural, & more politically “conservative”. When engaged by pollsters, they generally attempt to co-operate, which makes them comparatively susceptible to pollsters’ carefully constructed questions. This makes it easy for a hired pollster to accumulate the numbers they want to deliver
Cell culture OTOH treats their phones entirely differently: as likely to ignore or block a caller as to answer (or more so), unique approaches to staying in touch, and no real sense of obligation to answer a call. They are far more disruptive to normal response accumulation, engaging with the pollsters in a challenging / questioning manner. This makes it easy for pollsters to eliminate them from the results.
in the poll in current question, the polling universe was 1200+; we don’t know if this is the number of contacts attempted, or of the accepted responses . This would inevitably involve a split between cell numbers & POTS numbers, and almost certainly result in more accepted conservative responses than theoretically ‘librul’ responses, due entirely to the phone culture involved.
To overcome this would require a structure intended to overcome it, such as making sure cell & POTS responses were *numerically* equivalent. Polling is very similar to telemarketing & direct mail marketing, in that there are lists: specifically, golden lists of numbers/respondents who consistently provide consistent responses (taking polls, GOP/Dem-friendly or hostile, falls for every sales pitch, buys every X, etc). Telemarketers for example LOVE lists of people who predictably spend money on telemarketed stuff - and they PAY TO HAVE THEM.
As I’ve said before, not all polling is even legit: a great deal of political polls are worked up by partisan pollsters-for-hire, seeking to provide their customers with numbers that support the contracted issues/positions. Such pollsters can easily obtain lists by region, by zip code, by state of those who will give the responses the customer wants, and those pollsters make bank
This is the reason some pollsters go out of their way to get a real cross-section of actual opinions in real numbers, and demographically broad; lots of times, they’re associated w/ universities & hired by media, interest groups, etc. Not that that makes them legit, but they probably don’t guarantee political results
Publishing that negative poll on aid for Ukraine gives the undecided reader psychological cover for choosing no.
In any case, polls should not factor into legislation, with a single exception: those polls that feature a ballot.
I find it unethical for TheHill to perpetuate this malarkey of presenting polls as if they’re news.
I might have aired such a complaint once or twice before.
(add) your insights into phone culture do much to explain systematic error/bias.