The Right Time for Love: Tracking the Seasonality of Relationship Formation

March 21, 2012 at 6:55am

Spring is just getting started, and that's a good chance for us to examine how relationships change with the seasons.

 

Using U.S. Facebook data from 2010 and 2011, we looked at how different times of the year affect the beginning and ending of relationships. We started by tabulating the changes from a non-coupled relationship status, like "Single" or "Divorced," to a coupled status, like "In a relationship" or "Engaged." We compared that figure against the number of changes in the other direction, from coupled to non-coupled, to calculate the net percentage change. As an example, 4% more people entered into coupledom in December 2011 than left it, a net gain for romance.

 

Admittedly, these numbers aren't exact. Some people might not change their relationship status to "Single" after a breakup. Instead, they might opt to hide it from their Facebook profile entirely. In these cases, we can't be entirely sure what happened. But considering the relative levels of coupling and splitting up across days, months and seasons still helps us understand the temporal patterns of relationship change among people on Facebook.

 

What Days Are People Getting Together?

 

The days around Valentine's Day and Christmas are good chances to try your luck or breathe a bit easier. The data showed far more people paired up around these times than joined the ranks of the newly single.

 

Feb. 14: 49% more new relationships than break-ups

Dec. 25: 34% more

Dec. 24: 28% more

Feb. 15: 22% more

 

Sometimes changes to relationship status are meant to be in good fun. The fifth biggest day for a net increase in relationships was April 1st, or April Fool's Day, which saw 20% more relationship initiations than splits. But unsurprisingly, many of these appear to be short-lived: April 2nd was the year's most extreme day in the other direction, with 11% more break-ups than new relationships.

 

What Months are People Getting Together?

 

When we take the daily data and group them into months, we can get a glimpse of how relationship changes vary seasonally.

 

[View Larger Image Here]

 

To create this chart, we partitioned users in the U.S. into three ranges based on their self-reported age on Facebook -- under 25 years old, between 25 and 44, and 45 and over. For each of these three age ranges, we calculated the net percentage change in relationships for every day in 2010 and 2011. We plotted the distribution of daily values, grouped by month, using box-and-whisker plots.

 

The boxes and whiskers represent the 5th, 25th, 50th, 75th and 95th percentiles of the net percentage change on the days in each month. The colors of each box represent the month's relative extremity in terms of relationship changes. We computed this by comparing each month's average with the overall yearly average, tinting a month green if it was above average in terms of net relationship gain and red if it was below average. The more the month differed from the yearly average, the deeper the hue.

 

Across age groups, the summer months are bad news for relationships. In 2010 and 2011, May through August were clearly lower than the other months of the year, suggesting the daily net change in relationships reaches a low during the summer. Interestingly, the large tail in February for the under-25 group can be explained by the fact that the net gain in relationships is highest on Valentine's Day.

 

What Days of the Week are People Getting Together?

 

We also found patterns over the course of the week.  In general, we saw a net gain in relationships after the weekend -- Sunday, Monday, and Tuesday were the biggest days for new romance. Relatively more break-ups happened leading up to the weekend, peaking on Friday and Saturday among people in the older age groups. Among younger people (those under 25), this weekly pattern was similar but temporally shifted just a bit, with the low point in net relationship growth coming on Thursday and Friday, followed by a slightly earlier weekend bump starting on Saturday.

 

[View Larger Image Here]

 

Why might these patterns exist?  One explanation could be that people looking for a change tend to end their old relationships in time to spend the weekend with friends or get back in the game with someone new. Or maybe, as the song goes, breaking up is hard to do, and having a difficult conversation has to wait until the work week winds down. As for the net gain in relationships at the start of a new week? That may be a visible echo of the weekend's festivities -- and the new social ties that result.

 

This report was prepared by Data Scientists Jackson Gorham and Andrew T. Fiore.