Greetings, my friends!
This is a quick (but important) Public Service Announcement about all of the recent articles (like this & this) claiming that Clinton leads in California — pay no attention to them! They are all based off of the same survey from an organization called “SurveyUSA.” I’m not going to mince my words — this survey is completely bogus and I will show you why.
Why the New Survey is a Crappy Survey
Okay, first off: what does the survey say?
SurveyUSA claims that, among 802 people who they believe are “likely to vote in the democratic primary,” 57% say that they will vote for Hillary clinton, compared to only 39% who say they will vote for Bernie Sanders — but wait! Before we all start despairing over this seemingly-insurmountable 18-point lead in the all-important California primary, let’s take a closer look at this so-called survey…
Now — what makes a good survey?
For starters, a good survey has a large sample size relative to the population — generally speaking, the larger the sample size, the more accurate the poll is.
Another very important thing to look for in a survey is whether the demographics of the survey (at least, somewhat closely) match the demographics of the election that the survey is trying to predict. For instance, imagine I’m taking a poll to see how many Californians enjoy fast-food. In one poll, I approach a thousand random people on the street but, in another poll, I approach a thousand random people that are in line at McDonald’s — the results of my two polls would probably be different, right?
Now, let’s take a look at this poll that all the news stations are raving about… (for those of you who would like to follow along, the poll can be found HERE.)
In the section where it says that 57% are voting Clinton and 39% are voting Sanders, you’ll notice that the sample size is 802 “actual & likely democratic presidential primary voters.” In their chart, you can also see that they’ve broken down their sample based on age, race, ideology, etc.
Disasterous Demographic Disparities
The first red flag is the section where it breaks down the age of the sample.Notice that a whopping 42% of their sample is over 50 years old! Now, remember what I said about matching the demographics of the poll to the demographics of the voters — do you really think that 42% of California voters are over 50? Take another look — 17% are over 65!
Don’t worry — I’ve saved you the trouble of having to research California’s age demographics (my source is here, though)!
According to the census data, the median age in California is 33.3 and about 29% of Californians who are over 18 are 50+. This means that, in this survey,there are 13% more people aged 50+ and 13% less people under 50 than there actually are among eligible voters in California. In other words, the age demographics are, at the very least, skewed by about 26%. In reality, this distortion is probably much worse because of California’s YUGE, recent surge in new voter-registrations — and guess what? 2/3rds of them were under 35.
So — if the age of the sample size was skewed heavily toward an older demographic,which candidate’s percentage-points would go up? I’ll let you figure that one out.
Now, let’s talk about income-levels. If you scroll to the right, you’ll find the breakdown of the sample by income — notice anything… hmm, I don’t know — odd?
43% of their sample makes more than $80,000 per year! Holy Cow! If you look over at the census data, you’ll see that about 75% of Californians make less than $80k/year — which means that, compared to the general population of California, there are 32% more people who earn $80k+/year and 32% less people who make under $80k.
That’s a serious distortion! Basically, what I’m trying to say is that this was a survey of mostly wealthy, old people.
Messy Methodology Mistakes!
Okay — just in case the 26%+ distortion of age and the 64% distortion of income demographics weren’t enough, already — I’d like to shine a big, bright spotlight on the dubious methodology of the survey.
You’ll notice that only 63% of the sample were surveyed by phone (which is the standard, more-reliable method) — so, how were the others surveyed? After carefully examining the vague statement on their methodology, it seems that the remaining 37% were surveyed online! (For those of you who are unfamiliar with polling methodology, online survey techniques are generally frowned upon and mostly disregarded as uselessly inaccurate)
Their general statement of methodology (which you can read here), along with the more-specific statement that they’ve included with the poll itself, seems to admit to using some kind of online questionnaire for the participants that weren’t reachable by land-line. The only description of their process says, “38%… were shown a questionnaire on their smartphone, tablet or other electronic device.”
Do you even poll, bro?
Although the art of polling is nuanced and ever-changing, there is no serious pollster (that I’m aware of) that regards online-questionnaires as an accurate method. In fact, those who value their reputations steer clear of them, completely.
Disregarding the online-questionnaires, 63% of 802 is only 373 people surveyed and — with an estimated 8 million people voting in the democratic presidential primaries — the margin of error increases to 5%. But that 5% margin of error would only be accurate if the survey’s sample were similar to the overall demographics of the election.
You are being lied to
In addition to the fact that this survey was drawn from a sample that was extremely heavy with wealthy, older people (a demographic that favors Clinton), it was also a much smaller sample size than you’ve been led to believe — well, a much smaller accurate sample size, anyway. Oh — and, if you look up the four TV stations that sponsored the study, you’ll notice that those companies show up on Hillary Clinton’s donor list. So, yeah.
My advice is to completely ignore this crappy survey — it was designed by a corrupt media to distort your perception of the elections. As I’ve warned you in previous posts — they want you to give up hope. Don’t let them mess with your head — check their sources, think critically about what the news tells you, and use math! They hate it when you use math!
Good luck, my friends — keep your heads up and bern your TVs!
I’ll write to you all again, soon.
John Laurits #SeeYouInPhilly