With the numbers debate becoming increasingly involved, and now that it is clear that we RI bloggers are no longer merely talking among ourselves, I thought to expand my examination of population and wealth trends in Rhode Island. The most straightforward method is to start where I began with my proclamations of middle-to-upper-class flight from the state:

As I’ve explained before, this chart shows that, from 2005 to 2006, Rhode Island lost about 30,000 people whose households had been over three times the poverty level (roughly $60,000 for a family of four), and almost 22,000 over twice poverty. As a matter of personal history, I made limited use of this information until, at the tail end of December, WPRO’s Matt Allen spurred me to look again, at which point I was actually surprised at how high the number had been.
Of particular concern, as I began to broaden my inquiry, was the observation that, stepping back a year, from 2004 to 2005, the number over five times poverty actually grew 22,058, over three times poverty 21,754, and over two times poverty 21,166. In other words, 2005 saw a spike at the high end, and although the 2004–2006 results still show losses of 2,498 (5x), 6,495 (3x), and 471 (2x), I have to admit that I had begun charging as if I carried a sharper sword than I actually held.
Unfortunately, 2004 was the first year for which the Census broke out the income to poverty ratio with such a degree of granularity. However, the income to poverty ratio is provided up to twice poverty beginning in 2002, with the following results:

Whether it has something to do with the measurement of the statistics or actual events in the state, I haven’t been able to determine. It is clear, though, that the spike in 2005 was more the oddity than the plunge in 2006. Trying to explain the data, I turned to U.S. Census household income data.

As is plain, the spike was specifically among households earning between $100,000 and $199,999 in the twelve months before the 2005 survey. For a view that highlights the range on which we’ve been focusing thus far, the following chart zooms in on households with incomes over $50,000.

By the numbers, households with incomes over $75,000 per year (which one can very roughly compare with three times poverty) increased 12,364 from 2004 to 2005 and 5,032 from 2005 to 2006. Households with incomes over $100,000 per year (which one can somewhat more accurately compare with five times poverty) increased 14,092 from 2004 to 2005 and 23 from 2005 to 2006. Although both groups advanced more slowly in 2006 than 2005, all of the numbers are still positive.
However, factoring in household size helps align this data with the income to poverty ratio data above: From 2004 to 2005, married households earning over $100,000 per year increased 11,048, but from 2005 to 2006, they decreased 815. All of that decrease came from married with kids households, which only increased 6,007 in 2005 and which decreased 1,033 in 2006. What I’m getting at is that the decrease on the household income chart came in large part from families, whereas the income to poverty ratio data is individuals whose households have that much income. That doesn’t explain all of the difference, but it does describe one ripple in the wave.
Because the clincher of this post (as long in coming as it’s already been) comes via IRS data, let me offer the related figures from that data set. Perhaps the most significant finding is the support that the comparison lends to the hypothesis that families are partly (maybe mostly) responsible for erraticism of the income and income to poverty ratio trends. Lower income households are less numerous than lower income returns, while the opposite is true at higher incomes, which would happen if couples filed separately or households had a third income earner.
Don’t forget, when comparing these to the Census charts, that the IRS has not yet released data for 2006, so the downturn is not present in the data.

And now, without further delay, here — purchased with a small percentage of readers’ generous donations over the past couple of weeks — is the data that ought to simply end the conversation about what’s happening and propel the conversation about what we ought to do about it.

Lest disbelief lead you to doubt your reading ability, the chart does indeed show net losses in tax returns of 1,818 from 2003 to 2004, 3,869 from 2004 to 2005, and 4,427 from 2005 to 2006. The corresponding loss in citizens’ adjusted gross income over that total period was indeed $513,117,000, and to the surprise of nobody who’s read the above, 2005 marked a huge increase in that amount.
One might surmise, although my research hasn’t gotten far enough to state it as more than a theory, that the leap in “wealthy” citizens in 2005 was the result of houses sold in the course of leaving the state. The migration data is attributed to the year in which the return was filed (i.e., the year after the income in question), while the Census data is income over the previous twelve months for a survey that took place throughout the year, so both would be in keeping with a large burst of income at some point during 2004 and 2005 and tax returns filed in 2005 and 2006.
As one final piece of the puzzle (for now), I compared the average AGIs of those coming to the state from and leaving it for the five counties in Massachusetts and Connecticut that are right across the border. These populations, it seems to me, are the most likely to include migrants who either work in or have close ties to the areas that they are leaving. They might include, among other groups, low-income families pursuing Rhode Island’s extended benefits (incoming) and workers who just can’t take the taxes and poor infrastructure of their home state (outgoing).

I suppose the shadow of doubt is in the eye of the beholder, but it would take quite a bit of argument to persuade me that middle-class and above citizens of Rhode Island — especially families — are not loudly declaring — by placing their families beyond the tax collector’s reach — that they will not accept the likely solutions of the General Assembly to our financial crisis.
Now what do we do about it?
ADDENDUM:
One factor on the table that I didn’t address above is the Phoenix Marketing data that has Rhode Island climbing from the 20th highest percentage of millionaires to the 17th. Thus far, I’ve mostly pointed out that our neighbors, Connecticut and Massachusetts, are consistently in the top 5, and some have suggested that we can’t make that comparison. However, upon closer inspection, it turns out that Rhode Island’s “improvement” was largely due to the number of non-millionaire households that have left. (We also have a discouraging foretaste of the likely results of Census and IRS data collection for 2006 and 2007.)

If Rhode Island had not lost 1.1% of its households from 2006 to 2007 — that is, if its household population had remained the same — then millionaire households would have accounted for 5.26% of the total, which would have put us right back in slot #20, just keeping pace with national trends, perhaps because of some of the very policies that the progressives and the Democrats are heaving onto the chopping block.