"The Mobility Myth":

That's the title of an article in the April 2006 Reason (not online yet, apparently), and its premise is that, despite claims to the contrary, "Americans are more likely than ever to stay put." I looked up the census data that the article points to, and it seems to support the article's assertion: In 1947-48, the first year for which data was systematically gathered, 20.2% of all Americans had moved; in 2002-03, that percentage was 14.2%, the lowest it had been since 1947-48 (except for 2000-01, when it was also 14.2%).

My one concern is that the Census data shows an oddly large and sustained dip from 1999-2000 to 2000-01, which makes me wonder whether there that dip reflects a change in data gathering rather than in the underlying reality. Nonetheless, even if we set aside that 2% decline, the mobility rate would still be lower today than in the past.

If anyone has more data on this, or more information on the 1999-2000/2000-2001 dip, please post it in the comments.

James Lindgren (mail):
I'm not sure, but if I remember correctly, there were large employment drops in that period that were caused by reweighting ethnicity (1999-2001). So I would guess that any sudden drop in the data you mention is caused by reweighting adjustments based on the ethnic differences in the newer census. But that's just a guess.
3.20.2006 2:22am
The census data you linked to suggests that the drop in the number of people who moved occurred mostly within the "same county" category. The large drop, which may reflect a change in methodology in that category occurred after 97-98, and again after 99/00. The first dip was accompanied by marginal increases in other categories, so it's not obvious in the total numbers. The second dip is accompanied by decreases in those other categories.

The people who moved from one state to another (real mobility) appears to have dropped by maybe 10% since 1950, but may be slightly above levels of the mid-90s.
3.20.2006 2:26am
A couple more thoughts:

1) Depending on how the categories are correlated, the first dip after 97/98 may just be moving people around in the various categories - but there's obvious correlation in the data by eye.

2) Regardless, the state-to-state movers (the relevant aspect for the mobility myth, unless one wants to look at something like the effect of housing prices) should be compared with the number of young people, since old people move much less. Since there are presumably many more elderly folks now than in 1960, a 10% decrease in the number of state-to-state movers might actually suggest that younger people are as or more mobile as they were in 1950.

Without having seen the Reason article, I suspect there's a bit too much fiber in the author's diet.
3.20.2006 2:56am
Freddy Hill (mail):
This is not surprising to me. Telecommunications technology allows more and more people to stay put even when they are promoted, when their jobs change or even when moving from one company to another. I rutinely participate in video or audio conferences, IM chats or webcasts where people working on the same long-term project from all over the US are connected, often working from their homes. Large corporations such as IBM or Bank of America, to name just two, go to great lengths to avoid moving people around.
3.20.2006 2:56am
Kevin L. Connors (mail) (www):
I think the dip reflects the pop of the Clinton Bubble.
3.20.2006 4:50am
meep (mail) (www):
It's interesting Freddy Hill should mention it, but I come from an IBM family in that I have a lot of relatives who used to work or still work for IBM. In the 70s and 80s, IBM would relocate people at the drop of a hat -- but amongst everyone I know who works at IBM now, pretty much everybody stays put. Then there's the issue of "company loyalty" - my own company tried relocating a bunch of people from NYC to Charlotte, NC and lots of people decided they'd rather go work at a different company. Nobody felt like they had to stay with a particular company until retirement, unlike my dad's day.
3.20.2006 5:58am
Kovarsky (mail):
I wonder if there's a cross-section of the data that could control for telecommunuting, although I'm not sure how you would do that. You might be able to isolate particular industries, but then you're introducing all sorts of sample set bias based on industry cycles, etc.

I'd also add that not only has telecommuting allowed people to stay put, but it's also dragged the price of real commuting way down, so more people can afford to meatspace-commute without moving residences.
3.20.2006 6:49am
Nobody Special:
They used 1947-1948 as their comparison year? The year when a huge load of GIs came home and began college on the GI bill?
3.20.2006 8:26am
It makes sense to me that if home ownership is at an all-time high, mobility would be at a corresponding low -- a relationship that I think is supported by guest1's observation.

Homeownership eliminates things like lease expirations (and I suspect eviction rates are higher than foreclosure rates, but that's intuitive, not fact-based), creates economic incentives to stay put, and leads people to identify more strongly with where they live -- thus requiring a bigger reason (like a job or another person out-of-county) to warrant a move...
3.20.2006 8:45am
Anon1ms (mail):
I wonder what effect two-wager-earner households have on mobility? I know in my case, I wouldn't consider a move unless the payoff more than compensated for the impact on my spouse's employment (both monetarily and self-satisfaction).
3.20.2006 9:29am
In Wisconsin, the fastest growing county in 2004 was Juneau Co. The reason? That was the year that the new prison opened. I was surprised to learn that inmates are included in Census data. Could the rising prison/jail population be factor in why the data shows people are moving less?
3.20.2006 10:09am
Eugene Volokh (www):
Nobody Special: Did you look at the Census table I linked to?
3.20.2006 10:26am
Jim Hu:
Ack..comment eaten. Links seem to be disallowed today (even though there is a helpful box to insert links... :^(

On my blog, I link to a different version of the article by the same author.

The problem is that "It depends on what you mean by mobility"
3.20.2006 10:30am
Peter Wimsey:
I agree with guest1 that the really relevant numbers are the state to state moves; I don't think that movements within a county or within a state mean much more than someone is moving from one suburb to the next.

But roughly eyeballing the state-to-state moves shows that the yearly percentage of people who moved state-to-state remained approximately the same from '47 to about '80 (around 3.3 or 3.4%), declined from '80 to '90 to about 3.0%, and declined further from '90 to present to about 2.7-2.8%.

It would be very interesting if we could tie this information to where people were moving.

Without more specific data, my inclination is to view the the more frequent '47-'80 moves as being related to the presence of lucrative factory jobs being located in the northern part of the county plus california (and the virtual absence of these jobs in the south and west). The first part of this period overlaps with some of the Black migration north, as well.

The decline of rust belt manufacturing after 1980 is well documented, as well as the establishment of manufacturing facilities in sunbelt states. To some extent this will cause some migration of its own (when GM's corvette assembly plant moved from St. Louis to Kentucky, virtually all of the 2,000 workers moved with the plant), but in general the decline of manufacturing jobs would reduce the incentive for migration for many people. (Although migration to Calif. woul still be popular).

I'm not sure how to account for the decline from '90 to present. I'm skeptical of the telecommuting theory, though, since I don't believe that there are enough individuals who, but for telecommuting, would have moved to another state. Most people I know who telecommute simply do it one or two days per week, to save on the commute into the office. I do know one person who remained employed with a particular company when the company moved out of state, but this individual would not have moved to follow the company...he was offered telecommuting because the company realized that this was the only way they could keep him as an employee.
3.20.2006 12:18pm
Don Miller (mail):
I will beg to differ that County to County moves are not significant when looking at mobility rates.

On the East Coast, and Mostly Northeast corridor, this may be true, but west of the Mississippi River, it isn't true. County to County moves within a State can show significant mobility. I live in a low population, large geographic area state (Idaho) (my county is larger than Rhode Island). It is reasonable to move over 400 miles and still live in Idaho. It is conceivable to move over 800 miles on a north-south move and still be in Idaho. Most Western states are large enough that in-state moves over 400 miles are common.

Rather than see County-to-County and State-to-State moves I would like to see a mobility study that compared moves under 100 miles and moves over 100miles. Distance is the key factor.
3.20.2006 1:10pm
Dudley Crawford (mail):
Maybe it's as simple as the Boomers getting older and not moving as much?
3.20.2006 1:12pm
Don Miller: county-to-county moves are less significant because the statistic includes both short moves and moves across a state. Some real mobility may be included in that figure, but the state-to-state moves will include less noise from short moves, and is more likely to be a significant statistic.
3.20.2006 1:31pm
Noah Snyder (mail):
Might the decline in the number of people in the military over that time period explain a lot of the decrease in mobility? I know my friend in the army moves a lot more than I do. And I'm stereotypically mobile (moved once for college, once for grad school, will move once again for postdoc, and once for tenure-track).
3.20.2006 2:28pm
"Mobility" is capacity to move. Looking at actual movement is only relevant if there's some correlation between ability and actual movement undertaken. The "myth" may well lie in sentiment (how easy it seems to move), rather than results (how many move).
3.20.2006 4:04pm
nrein1 (mail):
I agree with the commentators who point at important factors as potential answers, the millitary being smaller and young people (those who are more likely to move) making up a smaller percentage of the population as being possible explanations. One other thing I have noticed among my firends, Ia m in my mid 20s, I know a lot of people who find it more convenient to keep their permanent address at their parent's house due to the large number of moves they make. Also how do these figures take into account college students. It seems to me with more people going to school now, that would mean more people moving though as I noted above they may not officially change their address.
3.20.2006 4:08pm
Jason Fliegel (mail):
The real question (because this is a legal blog after all) -- does this statistic put lie to the federalist argument that if people don't like the law in one state, they'll simply move to another, and the market of states will work it all out so that pro-choice people will move to one state, and pro-life people will move to another, and everyone will be happy? Or is decreased mobility a symptom of the increased homogeniety of laws across states due to the expanded scope of federal power and the growth of uniform codes of law?
3.20.2006 4:44pm
ElamBend (mail):
The census data can't tell us, but I wonder if their is a cohort of people that are constant movers. If I took my family back five generations, I would have four states, on either side of the family. I can think of only one move that was job related.
Except for me and my siblings, it has been from one rural area to another. It wasn't until I was older that I realized that most people's families don't follow this "wild-geese" pattern.
3.20.2006 7:38pm