Red Versus Blue US States

Who watches more porn – Republicans or Democrats? That’s the latest question our friends at Buzzfeed posed to us. Previously, Buzzfeed asked us which states watch more gay porn – and the results were quite interesting.

This time around, our Pornhub Statisticians set out to answer their question by first looking at the viewing habits of all the states across the US of A. Once we knew the average number of pages viewed per visit residents of each state, we compared that to the current population and political preferences.

Some interesting trends started to appear as they crunched the numbers, and that gave rise to another question – given the per capita pageviews for each state, was it possible to use porn habits to predict election results? Read on to find out!

Red Versus Blue

For our non-American readers, the color red traditionally represents the Republican party, and blue the Democrats. Looking at a map of the US, we can see that red states fill much of America’s ample mid-section, while blues tend to be concentrated in the West and North-East. The interactive map below shows the current political representation and the number of pages per visit for each state.

Starting with the total number of pages viewed per state, our statisticians divided that by each state’s population to get the pageviews per capita.
The average number of pages per visit are pretty close – 9.33 pages in blue states, and 9.2 pages in red states. But on a per capita basis, the gap starts to widen. Blue states view 137 pages per capita, whereas red states view 121 pages per capita – a 13% difference.

Red-Blue-Pageviews-Per-Capita

Have a look at the following two charts to see how each state stacks up compared to their red or blue neighbors. Kansas, with a population of 2.9 million and 8.9 pageviews tops both the republican and overall lists at 194 pageviews per capita. On the bottom end of both is Arkansas — with nearly the same 2.9 million population, they average 9.3 pages per visit, resulting in only 77 pages per capita.

red_states_per_capita3

On the democratic side, Nevada’s population of 1.8 million results in 166 pageviews per capita. Maine is home to 1.3 million folks, who typically go through 9 pages each time they visit pornhub, resulting in only 92 pageviews per capita.

blue_states_per_capita3

Porn, Pages and Predictions

Of the 10 states with the most pageviews per capita, only 2 are red states. Conversely, of the 10 states with the least porn viewing per capita, only 2 are blue states. If we were to just use porn consumption per capita as a predictor of political orientation, we could have predicted the results of the 2012 elections with 70% accuracy.

Using the same data, our statisticians could also have correctly predicted election results in 9 of the 12 swing states (75% accuracy). It’s these hotly contested battleground states that often decide federal elections due to their changing political preferences. Only Ohio, New Mexico and Wisconsin’s swingers managed to fool our porn per capita classifier. I guess we’ll just wait for CNN’s call in 2016.

Below you can see how our predictions lined up with the current political representation of each state.

red_blue_states_variation2

Conclusions

Assuming the porn consumption per capita is normally distributed for each state and that different states have independent distribution of porn consumption per capita, we can say with 99% confidence the hypothesis that the per capita porn consumption of democratic states is higher than the republican states.

Leave a Comment

  • SKB April 11, 2014, 4:46 pm

    Thank you so so much for providing this data!

    Reply
  • JohnnyMcNulty April 11, 2014, 7:05 pm

    What is the timeline on this? Are these visitor numbers per day, week, month?

    Reply
  • Pornhub Guy April 11, 2014, 8:14 pm

    Pages per visit.

    Reply
  • JohnnyMcNulty April 11, 2014, 8:23 pm

    I got that. But if there’s 92 page views per person in Maine, times 1.3 million, that’s 119,600,000 pages. Divided by 9.03 pages per visit, and that’s 13,244,739 visits. Just from Maine.

    Is that from the beginning of time?

    Reply
  • JohnnyMcNulty April 11, 2014, 8:23 pm

    and thank you for replying

    Reply
  • Matt Marcinkiewicz April 16, 2014, 8:07 am

    So ten visits per Maineiac (if I may) per year. Ok, I can see that. The data is probably skewed by people who are on there most days–that would go for all states, of course.

    Reply
  • JohnnyMcNulty April 11, 2014, 8:47 pm

    and I could be a complete dumbass, I understand. It’s just that the per capita page views are higher than the per visit page views in every state, which means that more people are visiting the site from each state than are in every state, which means it must be over a period of time.

    unless I am totally missing something.

    Reply
  • Pornhub Katie April 14, 2014, 4:16 pm

    The data is from the previous 12 months

    Reply
  • JohnnyMcNulty April 16, 2014, 1:48 pm

    Hey thanks so much! Cool.

    Reply
  • Crissa April 14, 2014, 9:10 pm

    I looks like you’re actually hitting a limit to the spread of broadband, when looking at the states independently.

    The states in the middle of the country have high broadband penetration because the earliest nodes on the internet were created using defunct oil pipelines. This lead to excess capacity which was easily tapped into. Those were also the states with high investment by the defense industry in the cold war…

    Too bad you can’t balance the numbers based upon internet penetration ^-^

    Reply
  • Matt Marcinkiewicz April 16, 2014, 8:15 am

    “Kansas, with a population of 2.9 million and 8.9 pageviews tops both the republican and overall lists at 194 pageviews per capita. On the bottom end of both is Arkansas — with nearly the same 2.9 million population, they average 9.3 pages per visit, resulting in only 77 pages per capita.”

    Lol. The author of the article obviously doesn’t understand the math behind the findings. Population and page views per visit are not enough to determine the page view per capita statistic. The crucial missing ingredient is total number of page views per state, which is not provided anywhere in this article, but which can be determined by multiplying the population figure by the page views per capita figure, then dividing by the pages per visit number.

    Reply
  • Roger Dodger May 3, 2014, 2:08 am

    Why assume that “porn consumption per capita is normally distributed for each state,
    etc.? There are many important variations among the states that would affect the amount and type of porn that’s consumed. Age, for instance. Income. Education. Household size. Urban/rural variations. These characteristics may well affect the political leanings of the states, but in different ways. Obviously, what we beed is a thorough (and expensive) multiple regression analysis to determine whether one
    state’s porn consumption leans more to one sort of sexual activity as opposed to another. Do Republicans in Blue states lean more towards straight, marital and
    Missionary position porn? Do Democrats in red states lean more towards masochistic porn — BDSM? How do the states compare, by party registration, as to multiple partners, oral, anal, manual, etc. These details would greatly improve the use of pornalysis in predicting elections.

    Reply
  • dat dere gains brah May 26, 2014, 10:11 pm

    I think you need that assumption to complete that hypothesis test. Otherwise, how can you claim that the difference in blue and red states is significant? They could use Central Limit Theorem to claim that even if the distributions are not normal, the distribution of mean porn consumption of red and blue states will be normal. But for CLT to work, you need at least 50 samples. We barely have 50 states in US.

    Reply
  • ken cook May 13, 2014, 4:28 am

    Assuming a normal distribution is incorrect, deviance like porn consumption has outliers, so all of this is bogus, but im glad you gave some amateur statisticians a job, lol.

    Reply
  • dat dere gains brah May 26, 2014, 10:00 pm

    Yeah, that’s not a good assumption. But I’m guessing that’s the only way they could arrive at the conclusion that blue states watch more porn. Without normality, it’s going to be difficult to complete that hypothesis test. And without the test, the whole experiment is meaningless as this difference could just be a random occurrence.

    Reply
  • dat dere gains brah May 26, 2014, 10:14 pm

    I think Central Limit Theorem could take care of it. Even if the distributions are not normal, the mean for red and blue states could be normally distributed if we use CLT. But it’s thin as you need at least 50 samples for CLT to work.

    Reply
  • Tim June 4, 2014, 10:35 pm

    The issue that I have with this page is that the numbers are shown in ‘per capita’.

    A better analysis would be ‘per broadband user’. The way the data is presented here skews the results towards blue states because rural areas have less access to high speed internet.

    Reply
Page 1 of 11