Crunching the Numbers on Love

I’ve spent the last few days crunching data from the largest-ever international survey of love.  Specifically, in 2006 and 2007, the Gallup World Poll went to 136 countries around the world and asked people, “Did you experience love for a lot of the day yesterday?” Betsey Stevenson and I report our initial analysis of the data in our latest column.  A snippet:

The good news: Ours is a loving world. On a typical day, about 70 percent of people worldwide reported a love-filled day. In the U.S., 81 percent felt love… Across the world as a whole, the widowed and divorced are the least likely to experience love. Married folks feel more of it than singles. People who live together out of wedlock report getting even more love than married spouses… If you’re young and not feeling all that loved this Valentine’s Day, don’t despair: You’re not alone. Young adults are among the least likely to experience love. It gets better with age, ultimately peaking in the mid-30s or mid-40s in most countries before fading again into the twilight years.

We’ve also compiled a Global Love Ranking. You’ll learn, among other things, that the Philippines is the most love-struck country of all, and on any given day, 93 percent of its citizens say they experienced love yesterday.  The US comes in at 26th, my native Australia is 44th, and the Armenians are dead last, with fewer than one-in-three experiencing love.

Here’s my Valentine’s Day Challenge: What explains the incidence of love around the world?  I’m interested in charts, regressions, narratives, and conjectures.  Either respond in the comments below, or play along on Twitter, using the hashtag #lovedata.  I’m sure we can rustle up some swag for the best analysis.

I should warn you, it’s harder than it looks. For instance, here’s my first attempt at explaining the data, simply using GDP per capita:

While the relationship in this chart is statistically significant, it explains very little of what’s going on.  I’m sure you can do better though, which is why I’ve published the full spreadsheet, complete with 3-letter international country codes (all the better for merging on new explanatory variables), in a Google Doc.

So, what explains love?  And just as importantly, what doesn’t?


Daniel

In order to understand what explains love, don't you first need to understand what love is?

This could be taken as a an unanswerable philosophical question of course, but for the purposes of this question it could be seen as a sociological one. Different societies inevitably have a different definition of love and indeed, the word itself is not ubiquitous across different languages.

Jack

I'm not a smart man, Jenny. But I do know what love is.

J

My sweetheart is Armenian, so I can only apologize to the other two people who must be unlucky in love.

DavC

Wouldn't it be more feasible that societies where people experience the greatest amount of love are the most prosperous thus having a greater GDP per capita? I mean, maybe love causes GDP and not the opposite

Renars

For me it seems that the graphical data represents a shape of heart...

Ryan

The heart (or U-shape?) could be significant: the very poor and very rich in terms of GDP less often have a low proportion of love. Just looking at the chart, the bottom 10 love proportions all seem to have non-extreme GDP. Perhaps the very "poor" have to look to love to remain content, while the very "rich" have more opportunities to be loved? How well does a parabola fit the data?

koda

lol, love is pretty simple. and generic numbers from population samples won't substitute for deep analytical thinking and development of hypothesis.

here's a tip, three roots of it are emotional dependence, insecurity, and the satisfaction of one's objective criteria in a mate.

koda

now i seem illiterate
hypotheses*

Pierre-Louis

I ran a few regressions using WDI data. It seems richer countries do feel more love... But an even stronger relationship is between income inequalities and love. The more unequal, the more love!
What I also found is that more love is correlated with lower military expenditures (as a share of total gov expenditures). Make love not war? And it seems love does not predict fertility... who would've thought!

Check out the scatters here: http://pierrelouisvezina.weebly.com/love.html

brazzy

Could be as simple as cultural differences in how seriously the word "love" is taken. A country in which it's unthinkably crass to say "OMG I just *love* Doritos" will probably score pretty low here even though the people there love and are loved no less than anyone else.

Enter your name...

It might be interesting to compare household size. More members means more opportunities for someone to have been loving. On the other hand, fewer sleep-destroying babies probably means less frazzled adults, so perhaps what's really wanted is the household size only of people over the age of three.

Eva H.

Falling below the line seems associated with factors like corruption, political upheaval, and trust (or delta-trust that's negative?).

According to a Pew article (Where Trust is High, Crime and Corruption are Low: Since Communism's Fall, Social Trust Has Fallen in Eastern Europe), "The survey also found that in Eastern Europe — a region where concerns about corruption are widespread — the tumultuous changes that followed the fall of communism have taken their toll on social trust. The percentage of Russians, Ukrainians and Bulgarians who believe most people are trustworthy has declined steeply since the early 1990s."

In comparison, many African countries in the Pew article perform surprisingly well in this love poll--perhaps because trust, while low, has been consistently low.

Being more emotionally reserved also seems predictive of falling below the line. Japan? Norway? Eastern Europe, again?

Read more...

Beth (@DataGeekB)

I ran the numbers and can tell you that there is no correlation between love index and the following demographic variables: total fertility rate, life expectancy at birth (male or female), difference between male and female life expectancy, infant mortality rate, crude birth rate, mean age at childbearing, median age of population, child dependency ratio, senior dependency ratio, population density, population size, population growth rate, or urban/rural distribution of population.

Source of demographic data: United Nations (http://esa.un.org/unpd/wpp/Sorting-Tables/tab-sorting_population.htm)

The sociologist in me thinks it's cultural -- not economic or demographic -- conditions that lead to a loving society.

Alan T

Suggest you look into correlations with life expectancy, religiosity, and rates of marriage and divorce.

By the way, don't miss the charming Valentine's Day google doodle.

Kieron

I found a great correlation coefficient for your love data against Countries' Sex ratio for people over 65. (r): 0.26012262851943.

It's not surprising that having the a more balanced sex ratio for the 65+ means more love overall.

James

Didn't experience love today? What, you don't have a dog?

Jake

I wonder how Hungary is #5 but their neighbor, the Ukraine, is almost at the bottom. I've never been to either or studied either, but are things really that different?

JamesAbra

Is it me, or does that graph appear to heart-like?

TBOU

Do you have the column you used for the per capita GDP? It's nice to be able to recreate the data using the original points, instead of finding something that's close.

Nancy

This has to be April Fools' Day... How did you get the data to make a heart shaped scatter? I'm in love!