COLUMN: Statistics are deceiving
There has been much discussion in the Statesman about Utah’s dismal divorce rate of late. One letter on March 28th asserted that 52 percent of all Utah marriages end in divorce, compared with a national average of 49 percent). Divorce rates are among the most misunderstood of descriptive statistics, and people often make the mistake of simply noting that the number of persons who divorce in a given year is often around 50 percent the size of the number who marry in the same year. But this is an apples-oranges comparison since populations of two greatly different sizes are being compared. (There are many more married couples who are capable of divorcing than there are unmarried couples of marriageable age in any year.)
Other times, people compare the number of divorces in a year to the adult population as a whole and call that a “divorce rate.” But a rate measured in that way is very influenced by the age distribution of the population. For instance, if the percentage of very old persons increases in a population, the “divorce rate” (measured this way) will drop – even if the number of divorces remains constant – since older people are less likely to divorce. Since Utah has a relatively youthful population compared with the nation, this measure artificially inflates our “divorce rate”.
A more appropriate comparison to determine a divorce rate is the number of divorces that occur as a percentage of the number of married couples. After all, they are the only ones who can divorce. In 1999, the national divorce rate calculated this way stood at 1.9, while Utah’s rate was only 1.3. That is, while 1.9 percent of all married couples in the nation divorced in 1999, only 1.3 percent of Utah’s married couples did. By this measure, Utah’s divorce rate is not so dismal at all, being only about two-thirds the rate for the country as a whole. And this low rate is still somewhat inflated, since Utah has a higher percentage of married couples who are in the early years of marriage – when divorce is most likely. An age-adjusted comparison would favor Utah even more.
Finally, the letter of the March 28 claimed that the lifetime likelihood of a marriage ending in divorce is 52 percent, compared with a national likelihood of 49 percent. As you can anticipate from what I’ve written above, a projection of Utah’s lifetime probability of divorce could not be higher or even as high as the nation’s, since the rate of divorce within the married population is so much lower in Utah. The writer’s source was misinformed. Assuming that current rates of divorce remain constant (which is how such projections are made), the nation’s lifetime probability of divorce (for persons who married in 1999) would indeed be approximately 49 percent, but the likelihood that a Utah marriage contracted in 1999 would end in divorce would be only 34 percent. Again, Utah’s marriages are much more likely to survive until the death of one or both partners. In fact, Utah marriages are about 1.3 times as likely not to end in divorce.