The States That College Graduates Are Most Likely to Leave | Hacker News Reader

The States That College Graduates Are Most Likely to Leave

nytimes.com

99 pointsPNWChris68 comments

Not sure about the methodology here. Say you're born in KY, but move to MA for college. And from there you move back to KY after graduation. The first move doesn't count, since you're not college-educated yet. But the second move counts as moving from MA to KY? Is that why MA is losing college graduates, as per this study? Because MA is chock-full of colleges, thats for sure.

Right underneath the article's chart:

>Note: Those who grew up in one state, went to college in another, and then moved again are counted as migrating from the state where they attended college.

Also, I'd imagine that this methodology doesn't affect things too much, considering that around 70% of high school graduates go to colleges in-state.

> I'd imagine that this methodology doesn't affect things too much, considering that around 70% of high school graduates go to colleges in-state.

So a full 30% go to colleges out of state? And I would wager that quite a few come to be close to family afterwards. That's a huge swing in the data.

Especially as people who have left their state to go to college have already left their state to advance their lives, regardless of what they do later.

So in the KY->MA->KY example that would count against MA and not KY, no?

I don't think that example counts against anyone. My reading is that only KY -> MA -> NY counts against MA instead of KY.

"People who grew up in one state, went to college in another and then moved somewhere else are counted as migrating from the state where they attended college"

To me "somewhere else" means a third state neither KY or MA.

Correct. They expressly call that out as being a disadvantage for MA when pointing out certain types of locations are doing better on average with a few exceptions.

I wish they had fixed it, rather than merely acknowledging it...

IMO this version is useful data, but it would be more useful if they also showed other variations. How many states lose kids that go out of state FOR college, what % of people who go out of state for college then come back after? (made up example) Say Maine has a ton of people go out of state to get educated but then bring all that knowledge back home, maybe that says people from Maine are dedicated to their home but don't think the education is good enough there to let them do what they want.

Each of those is an interesting data point with different ramifications.

Nobody comes home to Maine if they leave. At least not for ten or twenty years until they have dug themselves out of the hole.

source: Native Mainer, who moved to New Hampshire to get a decent job, who would love to go back home if the whole remote IT job revolution ever actually happens.

I sort of thought that might be the case but it was the first thing other than my own home state to come to mind because my Dad is planning a trip up there for next year.

The methodology makes the whole study suspect. New York, Massachusetts, Michigan, and Pennsylvania have lots of universities that attract out of state students. They get dinged for out of state students who go to college there and leave the state for work, but get no credit for out of state students who come for college and stay.

It's possible that the overall effect of out of state students is small but we have no way of verifying that with the data presented.

I'm curious as to why you think it's such a major issue. The question asked is not "which states are the best at attracting people to their universities", the question asked is "which states are worst at keeping a student as a longer-term resident once the student has completed university in that state". And to answer that question you must report the results the way they have been reported here.

The issue really seems to be that you want a different question asked. In which case feel free to do the research to answer your question!

The content of the article makes it seem like the question being answered is "what states are the best at attracting college graduates?" The actual answered question, which you state above, is an artifact of the limitations of their data.

Say that NY and NJ are equally attractive destinations for college graduates. But say NJ has no colleges, and everyone goes to school in NY. Under the methodology of the article, NY will show a huge out-migration and NJ will show a huge in-migration, even though they are identical in terms of keeping long-time residents who graduate college.

I think that phenomenon explains exactly what you're seeing with pairs of neighboring states. Is anyone going to seriously argue that NJ or CT are more attractive destinations for college graduates than NY or MA?

In order to determine which state is "worse" at keeping students, you need to know how many students were predisposed to leave.

State A has 10k students, 5k from out of state, and 3k leave.

State B has 10k students, 1k from out of state, and 2k leave.

Which state is better at keeping students?

I think that the only question this study answers is "is a follow-up study with a better methodology required" and the answer is yes.

This study points out an interesting sociological dynamics we do not understand fully (pardon my ignorance here as I did no verify if there actually exists a better study) and looking further into it would only create more confusion and possible false assumptions.

There are plenty of in state students of Michigan and Michigan State that leave the state after graduation.

I've gotten to know plenty of exceptional programmers, people with hackathon and open source chops and they all uniformly depart to Silicon Valley.

Silicon Valley companies and large VC firms are very organized and pursue them aggressively. First for internships and later for full-time positions. Then they get to the Bay area and despite what they think is a huge salary they soon learn that they need to pay half of it just to get an apartment.

Hey half your Michigan guys come to Chicago. Only half go to SV :)

You can begin assessing the size of the effect with data like this list of % out of state students at public universities: http://www.collegexpress.com/lists/list/percentage-of-out-of...

Anecdotal data point here. I grew up in upstate NY in a rural area. About 30% of my class joined the military and the remainder was split between college and work after school. At least 60% of my class no longer lives in NY.

That's not an atypical situation.

> The methodology makes the whole study suspect.

No it doesn't, it's a footnote that affects a few states significantly but has little impact on the bigger conclusion about inland states losing out to coastal areas.

The footnote doesn't provide the information that permits you to conclude it affects only a few states significantly.

Not sure why you would expect it to. The footnote clarifies how they classified people in one case that could have been counted in different ways.

Without that information, you cannot draw conclusions about the size of the effect of how they treat out of state students. And if you can't draw conclusions about the size of the confounding effect, you can't trust anything in the chart.

Yes, you have to draw on some common knowledge outside information to know that the size of the effect is not large.

Footnotes aren't there to handhold us all the way from set theory axioms to sociological conclusions. There are a host of potential confounding effects that could be large but common sense tells us probably aren't for various reasons. This is one of those.

This is not something you can hand wave away with "common sense." Common sense would suggest that college-educated young people are moving to Massachusetts (Boston) and New York (NYC), but the article shows the opposite. If the effect is clearly big enough to affect those states, what else does it affect?

E.g. why does Kentucky show in-migration and Kansas out-migration? Why does Alabama show out-migration while Louisiana shows in-migration? Arkansas shows in-migration but Missouri shows out-migration? Oklahoma shows in-migration but Kansas shows out-migration? College-educated people are moving to big cities in Wyoming but out of Nebraska? Are Connecticut and New Jersey showing a net in-migration of college-educated people because Connecticut and New Jersey residents are going to school in New York and moving back home?

shrug, you (the student) lived in Mass and then you left. Regardless of what state you went to after leaving, you are counted as having left Mass.

Regardless you are a graduate and the graph still reflects "Where Young College Graduates Tend to Move".

Yeah. It's unclear to me what's being shown here. Sure you can construct selective narratives. Young people are moving to Portland and Seattle even faster than California because they can't afford to live in California any longer and the cool urban west is where it's at!

But then you'd have to explain why South Carolina and Georgia have such high net immigration and why the Northeast generally is so negative. I suspect, as you say, that you need to factor in colleges explicitly.

The methodology is given but it makes me question what this show in a meaningful way.

Georgia is easy to explain. Well, more like Atlanta is easy to explain. No one cares about the rest of Georgia (I lived in middle Georgia for 4 years, wasn't pleasant)

Atlanta has cheap housing and good and growing jobs. Warm weather, few hours from the beach, few hours from the mountain. It has a little bit of everything. Tons of companies building new offices. Construction everywhere to accommodate jobs and people. Mercedes headquarter moved here from New Jersey. GM located a huge tech center, etc.

Now, on the map it has Florida with a net gain. I'm from Florida, went to college in Florida...left for work. I don't know what college grads are going to Florida for...lots of my classmates left Florida, and some actually came to Atlanta.

My assumption is that this is all based on how the US census / American community survey conducts its research. It tracks migratory patterns, but does not map individuals effectively.

There are a handful of datasets that track migration patterns, which can be used to assess movement in and out of geographic areas to other counties, states, and international territories.

The ACS does provide fairly granular geographic analysis, which could be used to look at individual colleges and their migratory patterns. If one were to care about the migration patterns of SUNY schools, for example, one could look to see whether those individuals stay in state, or then leave the state, upon graduation.

All in all, the ACS is a great tool for high-level research, but often opens more questions than answers.

Yes, there's text below one of the charts that says: "Those who grew up in one state, went to college in another, and then moved again are counted as migrating from the state where they attended college."

"Note: Those who grew up in one state, went to college in another, and then moved again are counted as migrating from the state where they attended college."

Florida has the warm weather, no winter allure. Lots of people that I know that used to live in Chicago either moved to California or Florida, because the winters here suck and they want to escape them. I also know a couple that moved to Atlanta, and a few that moved to Arizona. But for all of them, it's because they got sick of winter.

Even for me, who's been trying to stay in Chicago as long as possible, it's the second biggest temptation for me leaving, outside of wanting better job prospects.

"if states are losing more college graduates than they are holding or bringing in, they’re effectively subsidizing other states’ skilled labor forces."

Defining those subsidies could be helpful, if this refers to the in-state tuition rates, then it would be interesting to see what proportion of students are in-state. Out of all the in-state students, how many actually move out of the state afterwards? I'm pretty confident that in-state students are more likely to stay in the state than other students. On the other hand, it would be very unlikely for let's say a student who grew up in Georgia and studied at University of Michigan to stay in Michigan after graduation.

This is also unfair in the methodology: "Those who grew up in one state, went to college in another, and then moved again are counted as migrating from the state where they attended college." They were out-of-state students, so didn't get subsidized.

If there's other subsidies (than the in-state deductions) which apply to the entire student body, then the argument below is as valid:

"if the US universities are losing more college graduates than they are holding or bringing in, they’re effectively subsidizing other countries’ skilled labor forces."

However, in that case there IS some "quick, easy solution."

I assume the quick easy solution to educating foreigners and then seeing them move home would be to award an unencumbered work visa on graduation?

Anecdata: That would have solved their problem entirely for some of my friends. I can easily think of a good handful of good engineers I know from college that were disappointed to have to return to their home country when they could not find work sponsors. That's not to say that there aren't also the ones who's plan from the beginning was to return home, but for others the US had become their home.

The common story that I heart was that employers were balking at the amount of risk and paperwork required to hire foreign employees.

The UK is going rather in the other direction with this: trying as hard as possible to ensure that foreign graduates leave the country when they graduate. :(

> it would be very unlikely for let's say a student who grew up in Georgia and studied at University of Michigan to stay in Michigan after graduation.

I would say it's not all that unlikely. There's professional and personal relationships (marriage, grad school, internships, etc.) that form over 4 years that can keep the person in state.

Well, we're talking about college here. Most of the people one goes to college with usually move to other places after graduation.

There's the Midwest Student Exchange Program between 7 states and other tuition reciprocity agreements that probably skew the data too.

http://msep.mhec.org/

> A study comparing thousands of American and British census records between 1850 and 1880 showed that nearly two-thirds of American men moved across county lines, while only a quarter of British men did.

Doesn't really seem like a meaningful comparison. The US is comparatively massive, with individual geographic and economic regions easily being larger than the whole of Britain.

Not to mention parts of the US were literally still being settled during that time. The frontier didn't close until 1890.

In a sense US is still being settled. Populations are shifting and adjusting.

The United Kingdom has had two hundred years to adapt to the effects of industrialization. The United States got in on that much later, with the majority of it happening post-WWI, and at that point the modern layout of the country had only barely been realized. As you point out the frontier was still very much a thing in the years immediately prior to that.

There's still a lot of shifting within the US that you simply don't see in other countries. People are highly mobile, and cities like Austin and Atlanta have changed dramatically in character because of large-scale migration.

The only country actively changing more than the US is probably China.

The UK has had only about 30 years to adapt to de-industrialisation, or at least the end of mass industrial employment. There are definitely places hollowed out by it.

The US has those same places that have been hollowed out by de-industrialization. Take Detroit, for example. But there are plenty of other examples.

Can anyone from New England comment on why Vermont and New Hampshire are so different? Demographics and weather are similar, so surprised to see that Vermont is highly negative and New Hampshire highly positive.

I assume it's related to the fact that New Hampshire has more industry than Vermont (concentrated near the Mass border).

But, honestly, as I said in another comment, I'm not sure how much these numbers mean. With respect to to the MA negative numbers, if you've been anywhere near the Boston/Cambridge area recently the idea that there's this mass net exodus of people is pretty silly. Certainly if you've looked at Boston Metro real estate prices or the construction going on in the Seaport or around Kendall Square.

EDIT: It may also be related to MA people living in NH for tax reasons.

> mass net exodus of people is pretty silly

Well not necessarily "mass" (pun intended?) but -2% from 2000 to 2015. Note that everyone who moved from other cities in Massachusetts to Boston (and Boston burbs) are considered neutral by this metric. And there's always bound to be graduates of those Boston universities headed for the other coast. Also note that things are very different in Boston and elsewhere in 2016 from how they were in 2000.

Southern New Hampshire is within commuting distance of Boston, and has reasonable variety of the type of jobs you'd expect to see about a hour or two away from a major metropolitan area--also there are still some farmland being converted to residences which makes housing relatively affordable compared to Greater Boston. These forces fall off once you move beyond the southern NH cities--Northern New Hampshire is a lot like Vermont, but with less tourism.

It's probably almost entirely that south-east New Hampshire is a bedroom community for Boston. Most of Vermont is far enough away from Boston and New York that people can't easily commute, whereas if you live in Portsmouth or Exeter or Nashua, NH, it's under an hour to Boston.

I wouldn't really describe SE NH as a bedroom community for Boston though there's some of that. A lot of Massachusetts tech, for example, is actually North and Northwest of Boston for various historical reasons. There's also a fair bit of tech and other industry in SE New Hampshire. But your basic point stands--that SE NH (in particular) has access to tech and other good industrial jobs that Vermont residents generally don't.

As someone who worked used to work in tech in Woburn MA (30 minutes North of boston), and had a fair number of NH living co-workers, I would say this is true. Though a lot of tech is moving into the city proper as commutes here continue to worsen.

> People who grew up in one state, went to college in another and then moved somewhere else are counted as migrating from the state where they attended college.

I suspect that may have something to do with it. Maybe Vermont has more initially out-of-state college students, or they represent a higher relative proportion? Also, keep in mind that NH is 2x the population of VT and is close to a major metro area (Boston, ~1hr away). I have friends that live in Southern NH and commute to northern regions of Boston for work.

I'd like to see a 20-year later study that looks into the number of people that went to VT for school (from out of state), went to another state for a career (not a huge number of options in VT) and returned late. Everyone I know that went to school there wants to do this; they love the place but couldn't find a non-service/tourism job post-graduation.

May be NH people go to MA to college then head back to NH before they're 40?

I'm going to guess the lack of a state income tax.

People go to VT for college but most don't stay.

Highly recommend Ed Glaeser's book if you'd like to further understand the socioeconomics and urbanisation aspects of this trend: https://www.amazon.com/Triumph-City-Greatest-Invention-Healt...

I'd love to see the statistics for the state of NY outside of NYC.

I wonder why NY in general has a negative trend.

NY (along with a few other northeast cities) is somewhat unique in that of people who have kids and then move to the suburbs nearby, which I assume is a pretty significant demographic, a lot of them end up in other states like NJ and CT. But this same cohort pretty much anywhere in, say, California or Texas would not leave the state.

Plus there are a lot of schools in NY and a lot of people go to them from far away, which doesn't count in this graphic, but then a bunch of those people move back home after school which does count.

And there's a question of how they treat people who move around multiple times, some people probably move away for a few years then come back.

Overall because of all the confounding factors, this really isn't a very interesting graph. States seem like a pretty arbitrary split when thinking about this, it would make a lot more sense to be looking at cities.

The weather and the economy. Right after college, so many of has visited California, North Carolina, Florida or Colorado and stayed with people we went to high school or college with. In that mix is also NYC. So cities like Buffalo, Syracuse and Rochester all experience a "brain drain", and possibly ironically, as well as people leaving to go out of state, a number also go to NYC.

One theory is that New York's schools are more attractive to people looking to start college than the state's job market, conditions and amenities, and cost-of-living to retain those people after they graduate.

You can also work in New York City and live in two neighboring states, which changes the stats somewhat.

In my anecdotal, non-scientific experience, a lot of students who go to a university out of state may indeed leave that state upon graduation, but not often to return to their home state, often other states instead. So I don't see a big disadvantage to the methodology.

Does anybody have an idea why so many graduates are moving to Colorado?

Because it has been doing pretty well lately, economically speaking. People go where jobs are, and Denver is one of those places these days.

As to why that is... this is still a matter of debate, but here's one take. http://washingtonmonthly.com/magazine/novdec-2015/why-denver...

In my experience (of people from the east coast), a lot of people want to move to Boulder/Denver because they're "young" cities (and pot is legal).

Strange. It's an outlier on the map, and contradicts the author's main point. Yet there's no mention of Colorado in the article.

The States that college grads are most likely to leave?

Why, after recent political events, them United ones of A.

Reply on news.ycombinator.com