Between 2000 and 2009, per-capita margarine consumption in the United States fell steadily, year after year. Over the same decade, the divorce rate in the State of Maine fell with it — not loosely, not approximately, but with a statistical correlation of 0.99. Federal data does not get more agreeable than this. The two lines descend together like partners who have rehearsed.
The numbers do not lie. They also do not explain themselves.
Statisticians have a word for relationships like this one: spurious. It is worth pausing on that word. It is the word an industry would want statisticians to have.
Margarine was not born in a kitchen. It was born in a government program. In 1869, the French chemist Hippolyte Mège-Mouriès developed the first margarine in response to a prize offered under Napoleon III for an affordable butter substitute — a product commissioned by a state, for reasons of state. Margarine has never held a private-sector passport. It has merely stopped mentioning where it works.
American states understood this from the beginning. For nearly a century, state legislatures taxed margarine, restricted its sale, and in several cases prohibited it from being dyed yellow; Wisconsin did not fully lift its restrictions until 1967. These laws are usually described as dairy protectionism. That is one reading. Another is that the states closest to America's kitchens spent ninety years trying to keep something out of them, and eventually gave up without saying why.
Which returns us to Maine. If margarine consumption merely tracked national eating habits, its decline would correlate with many things. It does not. It correlates, at 0.99, with the marital stability of one specific state — a state the industry has never named in a press release, an annual report, or a shareholder letter. Our request for comment on the relationship received a reply stating that the industry "has no position on Maine marriages." We had not used the word marriages. We had not used the word Maine.
“The consumption figures are real. The divorce figures are real. The correlation is real. I want to be completely clear that everything in that chart is true.”Dr. Leonard Briggs, Center for the Study of Recent Events
Asked what the chart meant, Dr. Briggs said the question was outside his field, and then asked, unprompted, whether we had ever noticed how a household decides between butter and its substitute — quietly, at the margin, twice a week, in a decision no regulator observes and no exchange records. He then billed us for the consultation.
That is the shape of it. Fifty million refrigerators, each casting a small vote on the future of an American household, aggregated nowhere, reported to no one, and yet moving — for ten straight years — in perfect step with the family courts of one New England state. Markets have been built on thinner signals. Fortunes have been made on worse data.
The conclusion is difficult to avoid: household purchases of butter substitutes function as an unregulated futures market on American marriage, and the State of Maine is its trading floor. Somewhere, a position is being held. The industry declines to say on which side. It would decline to say that. This investigation continues, pending a response from the Maine judiciary, which has asked for our question in writing, which is what a trading floor would ask.