There are two demographics that media analysts have overlooked when searching for the source of Clinton’s unusual degree of support in West Virginia, Ohio, Pennsylvania and New Jersey. One is population growth and the other is the ratio of young voters to older voters. Lack of population growth is an indicator of economic stagnation, a factor that works in Clinton’s favor since her husband presided over a period of economic good times. The Obama phenomenon has attracted an unusually large number of new, often younger, voters. States that have relatively few of these younger voters will be less likely to support Obama. Both of these factors gave Clinton a strong advantage in the states being examined.
The states that have had the least population growth – growth below 5.0 percent – are North Dakota (0.5), West Virginia (0.8), Pennsylvania (3.4), Connecticut (3.6), Maine (3.8), Rhode Island (4.5) and Ohio (4.7). The states of most interest to us in this analysis are the rust belt states of West Virginia, Pennsylvania and Ohio. These are states where people have seen their jobs go overseas and their children leave to find opportunities elsewhere. The people in these states have seen change – change for the worse – and they are perhaps yearning for a return to the past. A familiar candidate whose husband presided over a time of prosperity would be more attractive than an unknown candidate who calls for change – a call that rust belt residents may hear as a move toward even more uncertainty. Many of these voters are blue collar workers.
The following states have a disproportionate number of people 65 and over as well as disproportionately fewer people in the 18-24 age group: Florida, West Virginia, Pennsylvania, New Jersey and Ohio. These states are not going to be fertile ground for Obama. He has shown a remarkable ability to inspire young people and to bring new voters into the process, but this appeal is not going to have as much impact in states with relatively few young people. On the other hand, the following states have an unusually high ratio of young to old; that is, the proportion of people in the 18-24 age group is almost equal to the proportion of those over 65: Georgia, Colorado, Virginia, Washington, Minnesota, Maryland, Kansas, Indiana, North Carolina, Michigan, and South Carolina. (Listed in order of ratios of young people.) Note that Obama did very well in almost all of these "younger" states. (There is no way to know how well Obama would have done in Michigan since he was not on the ballot,)
What are the implications of these factors for November? With Obama’s appeal to younger voters (and voters of all ages who have never voted before), many states that have not been in play for the Democrats in recent presidential elections are now in play: Georgia, Colorado, Virginia, Kansas, Indiana, North Carolina and South Carolina. With such a new and unusual phenomenon as Barack Obama, the electoral map of the past is of little use in estimating results in November. Looking at stale, static electoral maps from former elections ignores sea changes that are occurring in today’s electorate.
And just because New Jersey, Pennsylvania, Ohio and West Virginia preferred Clinton to Obama in the primaries does not mean that they will vote for McCain in the fall. Does anyone think that the people in these states that have suffered economic decline are going to turn to the Republican party? All Obama has to do is remind them that every time there is a Republican in the White House for 8 years or more, the economy goes bust. 1932 (Great Depression), 1992 (It’s the economy, stupid), 2008 (mortgage meltdown and jobs shipped overseas). (I did not include the Eisenhower years since he was not a doctrinaire Republican.)
In short, Obama should take New Jersey, Pennsylvania, Ohio and maybe even West Virginia in November because of the economy. And he will add a considerable number of states that have not gone Democratic for decades because of his appeal to new voters, independents and even some Republicans.
Note: I have not included the three largest states (New York, California and Texas) in this analysis since the populations of these states are so diverse that any one or two demographics are insufficient to represent the state. Also, I am not implying that population growth and age distribution are the only variables influencing the primary results. I am simply trying to introduce them as very important additional variables in trying to explain support for Clinton and Obama.
Data on population growth is based on the latest available census data; i.e., growth from 1990 to 2000. The data on age distributions in states is from a July 1, 2006 Bureau of Census report.