By now we are all familiar with the distinction between red states (where a majority of voters choose the Republican presidential candidate) and blue states (which go for the Democrat).

In the 2004 election, the final tally showed 31 red states and 20 blue states (including the District of Columbia). So, why were some states red, while others were blue? Did religion have anything to do with it?

To examine this issue, I consulted Barry Kosmin and Egon Mayer's American Religious Identification Survey (2001), a national sample of 50,281 American households in 48 states and the District of Columbia (Hawaii and Alaska were not included). The ARIS survey provides a detailed breakdown of religious identification by state.

The results point to five conclusions.

---First, the higher the percentage of the population that identifies as Protestant or Mormon, the greater the likelihood that the state voted Republican.

---Second, the higher the percentage of Catholics, the more likely the state was to be blue.

---Third, the larger the percentage of people who identify as Jews, the more likely the state was to vote Democrat.

---Fourth, the larger the percentage of a state's population that identifies with other religions or no religion, the more likely that state was to be blue.

---Fifth, when these findings are combined, the result is that states with the largest Protestant and Mormon populations were red, and the ones with the largest percentage of Catholics, Jews, others and "nones" were blue (see chart).

Of the 13 states that are 60 or more percent Protestant or Mormon, every one was red. Of the 28 states that are 40-59 percent Protestant or Mormon, 16 (57 percent) were red, and 12 (43 percent) were blue. Finally, of the eight states where Catholics, Jews, others and "nones" are the clear majority, seven (87.5 percent) were blue (such as California). Only one (12.5 percent) went red (New Mexico).

But, how does a state's religious composition affect its tendency to be red or blue? A complete answer to that question requires consideration of many factors that I cannot address in this space. However, at least one of them is the fact that religious identification affects the way people vote.

According to national exit polls, 59 percent of Protestants voted for the president. Unlike the last two elections, when Catholics favored the Democratic candidates, this year 52 percent of Catholics voted for Bush, while only 47 percent preferred Kerry. As in the past, Jews favored the Democrat, this time by a 74 percent to 25 percent margin. Also, as they have in the past, "others" and people with no religion favored the Democratic candidate (74 percent and 67 percent, respectively).

It is not known to what extent this national pattern was repeated in every state. For example, we do not know if Catholics in every state voted for Bush, or if there are some states where a majority of Catholics preferred Kerry.

So let us assume the national pattern occurred in most, but not all, states. If so, one reason every one of the mostly-Protestant and Mormon states went red is that the Protestant/Mormon majority voted Republican and got some help from others, including Catholics. States with greater mixtures of religious groups went red when Protestants and Mormons got significant help from others, especially Catholics in states such as Montana, Wyoming, Colorado, Arizona, Texas, Louisiana and Florida.

The states went blue when Protestants and Mormons did not get as much support from Catholics and others, such as in Illinois and Oregon. All but one of the states where Protestants and Mormons are a distinct minority went blue because Catholics, Jews, others and "nones" in these states tended to vote Democratic.

The religious composition of states certainly is not the only variable affecting the state-by-state outcome of the election. Other factors, such as the states' racial, ethnic and educational characteristics, also must be taken into account. But one cannot fully understand the results of the 2004 election unless one takes religion into account.

James D. Davidson is professor of sociology at Purdue University in West Lafayette, Indiana. His most recent book is Lay Ministers and Their Spiritual Practices (Our Sunday Visitor Books, 2003).