Demographic Fallacies

The New York Times had an article entitled “The Growing Blue-State Diaspora.”  In that article, the authors make a claim that as blue state residents move to traditionally red or swing states, they are changing the political ideological tendencies of their new home states.  It is not until almost the end of the article that the authors acknowledge that just because the person comes from a blue state it does not necessarily mean that person is politically liberal, or more apt to vote Democratic.

This is an important area of interest as political parties plot strategies for 2016 and beyond.  Most attention is focused on the Hispanic vote.  Other articles are highlighting the fiscal conservatism/social liberalism of today’s younger voters.  As for this latter category, most analysis tends to leave out the fact that young people of voting age- especially the 18-22 age group- have traditionally low turnout compared with other age groups and demographics.  Thus, although they may very well be fiscally conservative and socially liberal,  it does not infer that we are seeing a renaissance of libertarian thought.

Additionally, the analysis of this age group and the Times article also fails to take into account changing political ideological leanings over a life span.  Most younger voters do tend to be on the liberal side and grow more conservative as they get older, settle into jobs, start raising a family and become homeowners.  A young voter- a millenial- may start out with liberal tendencies, but that does not guarantee they will remain so all their life.  Thankfully for the GOP, they tend to stay home on Election any way.

Leaving aside the demographic group analysis, the article in the Times is about internal migration, specifically state-to-state migration.  According to the Census Bureau, the top reason for moving is better housing possibilities, followed by family reasons (change in marital status due to marriage, divorce or death), better employment opportunity and then “other reasons.”  It should be noted that “political reasons” is not an option under the Census Bureau analysis.  One could intuitively believe that the older the interstate mover, the less likely they are to change their political ideology or party affiliation.

It is very important to analyze where people are moving to and where they are moving from on a broad case basis.  In 2008, Bill Bishop published “The Big Sort,” a book about the clustering of politically like-minded people.  His studies were primarily focused on intrastate clustering since that is where the majority of internal migration occurs.  As he eloquently showed, it is not so much a conscious clustering of politically homogeneous people than it is mobility or the lack of mobility.  For example, it explains the white exodus from the major cities to the suburbs or neighboring counties.  It is a desire for better surroundings and especially housing opportunities that is the impetus for the move.  The question becomes whether the liberal city dweller carries their liberalism into the suburbs, OR do the suburbs temper the liberalism?  Granted, not all major city dwellers are liberal to begin, but they do show that tendency.

A good example is the Philadelphia area- the Nation’s 4th largest city.  The four immediate counties are Delaware, Montgomery, Chester and Bucks.  As Philadelphia’s population has declined relative to surrounding counties, we can assume that some former Philadelphia residents have moved to these suburban counties.  The next thing to look at is if any political ideological adjustments can be discerned by looking at the voting patterns in these counties over time assuming that the conservative voter will vote Republican and the more liberal voter will vote Democratic.

From 1960 to 1992, Democrats won Philadelphia by an average  of 32.9 percentage points.  As of 1996 through 2012, that average has risen to 58 percentage points.  Looking at neighboring Bucks County to the north, from 1960 to 1992, it favored the Republican Party by an average of 8.9 points, but after 1992, it favors the GOP by 2.2 points.  The same weakening trends for the GOP are seen in Chester, Delaware and Montgomery counties  where before 1996 the GOP held a 9-11 percentage lead.  Today, both Delaware and Montgomery counties have flipped to the Democratic side percentage margin wise.

The difference is Chester county which does not directly abut Philadelphia, but has Delaware county in between.  Before 1996 the Republican Party held an 11.4% average advantage; today it is an 8.9% advantage.  Since most of Pennsylvania’s intrastate migration is county-to-neighboring-county, the trend becomes demographically obvious.

As the upwardly mobile move out of Philadelphia itself, left behind are the more hard core Democratic/liberal voters.  The more moderate Democratic Philadelphia voters have moved to Bucks County while the more liberal Democratic former Philadelphia voters have moved to Montgomery and Delaware counties.  And while Chester county has seen a weakening of its Republican leaning base, it neighbors Montgomery and Delaware counties have shown a greater trend.  Hence, the more moderate Democratic voters are likely either bypassing Delaware and Montgomery counties, or the Philadelphia Democrat first moved to these counties then later relocated to Chester County.  This is evident because of the precinct breakdown of Chester County.  Those that border Montgomery and Delaware counties tend to vote more Democratic while those further removed from Philadelphia tend to vote Republican.

Most likely, the same can be said of any major metropolitan area in the United States.  For example, as the more affluent white Democrats moved out of Detroit, its Democratic advantage became more hard core.  Meanwhile, the more affluent and moderate Democrats that moved to the suburbs have changed the political landscape by bringing their Democratic tendencies.  In some areas, it has overcome the traditional Republican advantages.  Likewise, such is the case in Milwaukee and Los Angeles.  With the latter, the city’s population is growing, but that is attributable to a burgeoning Hispanic population.  As the middle class whites then move out (who trended towards the Democrats) into neighboring traditionally Republican areas like Orange County, we also see Republican advantages in that county decreasing over the years while the Democratic advantage increases in LA itself.

Certainly, the intrastate migration of people from metropolitan to suburban areas has had an effect on electoral politics.  Urban dwellers tend to vote Democratic.  By virtue of moving to a suburb does not automatically transform them into a Republican.  Where a shift in political ideology may occur is when the urban dweller moves to the more distant suburbs (or exurbs) of metropolitan areas.  However, there are two considerations.  First, a move across two counties is likely a more motivated mover to begin with and secondly, they are more likely to have been an urban Republican voter moving from the increasingly liberal urban center.

In fact, we see this in neighboring New Jersey.  Essex and Union counties in the north are Democratic strongholds that are declining in population.  They are losing population to the neighboring counties of Morris, Middlesex, Monmouth and Ocean.  Morris and Middlesex were Democratic-leaning counties before 1996, but have increased since then.  Monmouth is still Republican leaning, but certainly less so than before 1996.  The difference is Ocean County which is buffered from Union and Essex counties by Monmouth County.  Since we know that Essex/Union people are moving to Ocean County and since Ocean County’s Republican leanings have actually increased since 1996, the conservative/Republican voters are moving further away from Union and Essex counties.  Why?  The most likely explanation is that the political ideology of Ocean County appeals more to these people.

Hence, urban to suburban immediate neighboring county migration tends to modify the voting behavior of the neighboring county.  However, urban to more distant counties barely weakens the GOP advantages in the exurbs, or strengthens them.

If we assume the same dynamics for interstate migration, then the NYT article is fundamentally flawed.  Those willing to move a more distant location are more motivated and political ideology either has no effect, or a minimal effect.  Of course, there are exceptions.  For example, liberal East Coast migrants that move to Florida tend to cluster towards the south in the Miami suburbs.  We would expect them to make these areas more Democratic/liberal.  But, if they move to areas more distant from urban centers, they make the area either more Republican or more moderately Republican.  Likewise for the California/New York migrant moving to Texas.  It should also be noted that most interstate migration is to a neighboring state.  Hence, the net migration from New York to Florida or California to Texas has little effect since the numbers are so small. And consider the surrounding states.  Texas is surrounded by New Mexico, Oklahoma, Arkansas and Louisiana- all rather conservative states with the exception of New Mexico.  But, New Mexico- to- Texas migration is minimal.

Likewise, Florida is surrounded by conservative states- Georgia, South Carolina, Mississippi and Alabama.  Since the bulk of migration to Florida is from these surrounding states, one can predict that the influx of these migrants neuters the effect of the more distant migrant to Florida from traditionally liberal/blue states like New York.

The entire macro effect is a general evening or cancelling out of each other.  The practical effect is that certain congressional districts will change over time, but the statewide offices- Governor, Senate and President- will have little effect over time and the same battleground states in 2012 will likely be the same battleground states a generation from now.