Is wealth leaving the state?

Print More

The claim that wealth is leaving the state is often discussed, but does the data back it up?

We analyzed two decades of IRS statistics on income among people moving from state to state to see if we could answer that question.

Since the data are income-related, it doesn’t address accumulated wealth, but the income tax is the largest single source of state revenue, providing about $9.8 billion of the state’s $19.8 billion annual budget.

1. People moving out have more income as a group than people moving in

It’s virtually always the case that in a given year, people leaving the state take with them a higher combined income than people moving in, resulting in a net loss of adjusted income. In the data we analyzed, that trend held true in all but one year: 2003.

2. On average, people moving out don’t always have more income than people moving in

One reason more income leaves the state than moves in is simple: More taxpayers move out than in, according to the data.

When we look at the average – rather than the combined income – of people in each migration group, the results are far more mixed.

In some years people moving into the state had a higher average income than people moving out of the state; in other years they didn’t. This doesn’t refute the claim that wealthy people are leaving, but it doesn’t support it either.

It would have been preferable to have other data points in addition to average income, which can be skewed by extreme values. (We’d be happy to hear from readers what other data touch upon this topic.)

Here’s a look at the number of tax returns with change of addresses indicating a move out of Connecticut versus a move into the state.

3. Despite the out-migration, income usually rises

Looking just at movers misses the fact that most people don’t move in or out of the state from year to year. The combined income of new and existing residents increased in most years we looked at (15 out of 20), so it’s probably not fair to say there’s a net loss of income overall because of people moving.

While the number of people moving into the state has an impact on the net gains and losses from migration, it’s important to keep in mind that Connecticut has had the top adjusted gross income in the country, according to data going back to 1991, so the pool of people moving out is more likely to make more than people moving in.

About the data

The IRS statistics are based on the address changes on tax returns from 1992-1993 to 2012-2013. This data set has the benefit of combining the factors in which we’re interested – income and migration – as well as being pretty accurate, since it’s based on real tax filings. There are shortcomings, too. It leaves out people who don’t file tax returns, which could mean the poorest are left out of this data set altogether, skewing income figures higher.

The figures are based on 95 to 98 percent of returns and don’t include data from returns filed starting in “late September.” For more specific details on how the IRS compiles its state-to-state migration data, check out their explanation here.

Check our work

All of our calculations, including the code to transform more than 30 spreadsheets into usable data are available here. We invite you to check our work and use it as a starting point for your own research.

What do you think?

  • FDB

    Is there a way to test the possibility that a greater proportion of the people leaving CT (than of those arriving) are new retirees? This would make a difference to future earnings, which is of course what really matters.

    • Jake Kara

      Not from this data set, but that would be great to know.

      • FDB

        Maybe census data can shed some light on the relative ages of inbound vs. outbound movers. Might be interesting to see, even if those data can’t be linked directly to the IRS data.

    • SteveC1

      VERY good point!

    • StillInTheMiddle

      Census data details what you are looking for. In my town in Litchfield county we expect a 26% drop in school
      age population, combined with a 46% increase in the population over 65 (Source: 2010 Census, CT state data
      center). Not a pretty prospect for local revenues and I suspect it is likely representative of the majority of the towns in the State. It is pretty evident, as even the Governor is finally stating, the future is not so rosy and I think he is sugar coating it. We are at the financial cliff.

      • SteveC1

        Not all bad by any means – a 26% drop in school-age kids = decrease in town education dept. expense… that’s a pretty significant “plus.”

        • ED Doyle

          Steve Do you really think that the Ed Budget will go down by 26%? My guess maybe 2%.and with the mill rate increase your taxes will be the same.

    • Naro narosky

      Retirees can’t sell their homes. They are stuck in CT although they would love to leave.

  • StillInTheMiddle

    It would be of been more interesting to see the net income migration data prior to implementation of the income tax in 1991 to compare. Unfortunately the IRS is a year behind their mandate to post the data and has only posted data through 2013 so we are unable to see effect of the two recent record income tax increases. Considering the current budget woes one suspects the outflow trend curve is much steeper. Census data consistently confirms outflow with State population decreasing with only one county in 2015 growing and that was due to more births than deaths rather than positive net migration inflow.

    The IRS and Census data conclusively confirm why projected income tax revenue consistently lags Government projections. The continued loss of the more affluent tax payer, our educated young in search of financial opportunity and our minuscule State GDP growth coupled with excessive State Government financial commitments, both short and long term, it is hard to see how the current Administration and Legislature has the ability or political will to reverse the trend. I suspect they will continue to nibble around the edges while playing the blame the rich game. Those who are paying attention are voting, with their feet, as the data confirms.

  • Bud Morten

    Since, as you note, the data do not include “people who don’t file tax returns,” and since the State’s population is not declining each year by the gap between outmigrants and inmigrants, does this not imply that there are are thousands of inmigrants each year whose income is below the reporting thresholds (currently ~$10,000 to ~$23,000, depending on filing status)?

  • Bud Morten

    With regard to your title, “Is wealth leaving the state,” and your claim that “it’s probably not fair to say there’s a net loss of wealth overall because of people moving”: Since, as you say earlier, “the data is income-related [and] doesn’t address accumulated wealth,” why not use the title, “is income leaving the state,” and why do you then use the terms “income” and “wealth” interchangeably in your comments? Based on anecdotal evidence, it is likely that vast amounts of accumulated wealth are leaving the State to avoid its counter-productive estate and probate taxes. It is also likely that much of the accumulated wealth that is leaving is controlled by taxpayers whose incomes at the time they leave are well below their peaks because they have retired, thereby understating the significance of their outmigration in your analysis. The impact on the State of the loss of this accumulated wealth is substantial, two important examples of which are the loss of support for CT non-profits, and the loss of the human capital and civic leadership these people would otherwise provide to the State.

  • joshuavincent

    Good stuff, the data is from solid underpinnings, but wage income and wealth are two different things. Can wages serve as an stand-in for wealth? For poor people, yes, for middle-income people, probably. But for Fairfield , say, much wealth isn’t counted in wages: trust funds, land and buildings, assets, etc.

    If we find a way to measure wealth as opposed to wages, i suspect the outflow of wage income is also matched by sheltered wealth which can go anywhere as much as wages. With GE moving out, it is pretty much assured that Fairfield real estate and land values will go down, as the capital needed to prop up a property market will now be in the Boston Burbs.

  • fvcarstensen

    CT is a state that has seen essentially no job creation in thirty years (current employment is a gnat’s eyelash above that of February, 1989; the worst record by far in the nation), you might expect this dynamic to be worst that what this analysis shows. We do know that CT has been losing higher wage jobs and adding lower wages jobs. Indeed, CT began adding jobs in 2010, but its economy continued to shrink until 2012, having suffered a 17 quarter recession (from the end of 2007), the longest in the nation. And today employment is still below the previous peak, as is output. IF Connecticut had enjoyed even very moderate growth, we probably would not be so concerned about this–but Connecticut is in deep economic trouble and faces a trajectory that is deeply worrisome. The focus on the fiscal crisis is crowding out what is ultimately the more important challenge: how to restore even a modicum of growth, without which the fiscal crisis will indeed become permanent.

    The Legislature does have before it a bill that could unleash powerful economic growth and job creation: permitting use of stranded tax credits (nearly $2 billion) to fund major investments. That is that UTC is now doing, but it was a special deal; it did not involve adopting a general policy. Yet this approach is entirely self-funding: it generates so much net new activity that additional revenues more than cover the cost of cashing in the credits–see the UConn analysis, available at the CCEA website (March, 2010). If Connecticut does not adopt policies to “bend the curve,” wealth and income will be leaving the state, but the driver will be the continuing economic malaise.

  • Unless the data is adjusted for inflation and displayed in constant dollars, results over a 20-year period are not going to be indicative of any trend. $1000 in 1993 dollars has the same buying power as ~$1600 in 2013 dollars. And the “Tax returns moving in/out of Connecticut” ignores any underlying change in the total population of the state. It would be more meaningful to see the returns moving in and out as a percentage of the total population of the state.

  • meepbobeep

    It is awesome that you linked to your github with data + python code. Yippee! (and yes, I will be checking your work.)

  • MetroHtfd Prog Pts

    In addition to the point below about inflation adjustments for the data, you might want to look at the median return, not the average return. Since incomes will be skewed, the average income will be pulled up by the highest income returns. The median would give more sense of what a return in the ‘middle’ is like.

  • PureGenius

    Averages are great, but how about the range and distribution of incomes leaving and arriving? How many people in each category have an income that qualifies them to actually pay taxes, because isn’t that all that really matters?

  • John Doe

    I took a quick glance at both the inbound and outbound AGI for taxpayers in 2010-2011. It appeared CT exchanged it’s taxpayers with other state’s taxpayers that had significantly lower AGIs, in states such as TX, FL, MA, etc. However NY appears to be an anomaly. Shows about 7-8k of returns with outbound AGIs of 473k, and inbound AGIs of 938k. This small inbound population of high net income earners may be making CT’s numbers look better than actual.

  • bpphil

    This makes sense and it’s not a bad thing. Most of the folks moving out are retirees with savings leaving for places with warmer climates, like Florida. I would bet that most people moving in are doing so to build their lives. (This was me 10 years ago)

  • Naro narosky

    Real Estate prices in CT (outside Fairfield County) have collapsed to near 1980 levels. This is due to the outflow of citizens. The workers are replaced by illegal immigrants and the poor that strain our resources. The biggest tax of the CT residents is the depreciation of their realestate valuation which is continuing unabated.