• New and improved Krugman chart

    Some commenters (and emailers) suggested that I improve the chart I made that Paul Krugman posted by adding in state population with bubble size. Well, in the immortal words of Yul Brynner, “So it has been spoken, so it shall be done”:

    UPDATE: Man, you guys are picky. I redid the colors. Here’s the data:

    State CTR/Income Population
    Tennessee 0.222656052 6346105
    Idaho 0.198588989 1567582
    Nebraska 0.159104736 1826341
    Oklahoma 0.206684104 3751351
    Arkansas 0.244379332 2915918
    Louisiana 0.206262682 4533372
    Alabama 0.236314519 4779736
    Wyoming 0.139698511 563626
    Utah 0.143995998 2763885
    Mississippi 0.260953506 2967297
    Average Conservative 0.211749692 32015213
    Connecticut 0.146790688 3574097
    New Hampshire 0.150398409 1316470
    Rhode Island 0.210748799 1052567
    New Jersey 0.153863101 8791894
    California 0.163510428 37253956
    Hawaii 0.159355519 1360301
    New York 0.192638358 19378102
    Washington 0.170722276 6724540
    Oregon 0.205603994 3831074
    Massachusetts 0.165436292 6547629
    Average Liberal 0.170696378 89830630
    UNITED STATES 0.184604932 308143815
    Comments closed
    • OK, now label the states (or include a table).

    • That would get you a failing grade in any high school class. One axis was entirely unlabeled, the other had no explanation of the numbers. I have a masters that required me to take sophisticated statistics and quantitative economics, but I can’t figure it out.

      It’s a pity, since it looked like a very interesting chart.

      For those of us who aren’t immersed in the data every day, a few more details about what “current transfer receipts” are, might help, too. Adding a line or two of information many readers know, but some don’t, doesn’t harm those who know and adds to clarity for those who don’t.

    • Dear incidentaleconomist,
      This “problem” would go away if tax brackets were indexed by cost of living, and safety nets were more proportional to cost of living. Almost the entire trend can be explained by the presence of large urban areas, which are generally left-leaning and at the same time have higher cost of living/higher wages. On top of this, states without major urban areas are small states, and, because of the constitution and electoral college, receive disproportionately more political representation .

    • Though it’s fairly clear due to the size of the “United State” dot, it would make it easier to read quickly if the legend included the meaning of size. On the other hand, since size here just represents, uh, size, my opinion may be influenced by how many bubble charts I’ve read in the past…

      Anyway, nice chart!

    • Color of US is too close to color for conservative. And why not use red for conservative and blue for liberal? It is an expected convention. I’d use grey cross-hatch for US. Use blue cross-hatch for average blue and red for average red. Label the outlier states.

    • @SAO:

      Sure, Aaron does not make a perfectly explicit chart but I don’t understand why you are being deliberately obtuse when interpreting it. If you just click on the link, you’d know that the X-Axis is a measurement of conservative/liberal leaning on of the states. Furthermore, Paul notes that the graph shows the top 10 liberal and conservative states based on Gallup polling.

      For better or worse, blogging sometimes requires sorting through multiple links but for the curious – the information is there.

    • It’s interesting to see how quickly detractors of Aaron’s graph leap to alternative explanations or quibbles rather than honestly consider what this might mean if their explanations are shown to be false and if their quibbles are resolved.

      Social ills are concentrated in conservative states. In large part, this is because conservative states have a long history of discrimination against minorities and women that led to those groups having lower levels of education. Conservative states have also long been hostile to creative people. Creative people tend to flock together, which is why scientific strongholds like Boston and Philadelphia and San Francisco are also centers for the arts.

      It is no surprise that social ills are concentrated in conservative states. What is a surprise is that the rest of us bother to try to help them rise from their wretched state, especially since they tend to be so dishonest/hypocritical/downright unpleasant about receiving help.

    • So which transfer payments are we taking into account? Are agricultural subsidies in there, for example?

    • The title is misleading: the numerator data at the Tax Foundation are Federal spending by state. I have not found anything there labeled personal current transfer receipts, just “Federal Spending Received Per Dollar of Taxes Paid by State, 2005,” plus a similar one for the time period 1981-2005.

      • I didn’t use the Tax Foundation data. I used the BEA data that Paul Krugman references.

        Again, you’ll get a lot further with me by asking for sources that you don’t understand than by assuming I’m trying to mislead you.

    • There’s a lot of play in this data…

      First, the conservative states are in the south and thus typically warmer (AZ, TX, FL)… making them a good place to retire, and thus probably skews the results. Plenty of people work in the north, and collect social security/medicare benefits in the south.

      There’s also probably a big chunk spent on military contractors and bases… I have ZERO data to back this up, but I’d figure they would be more prominent in conservative states.

      A more interesting chart would remove Social Security and Medicare expenses before looking for ‘moochers.’

      • “A more interesting chart would remove Social Security and Medicare expenses before looking for ‘moochers.’”

        Why? The average SS/Medicare recipient takes out much more than he pays in. Isn’t that what’s meant by the term “moocher”? Why treat that case as qualitatively different from the working single mom receiving TANF benefits?

        • What data do you show suggesting that Social Security recipients get more than they paid? I only wish the contribution of my employer and me to SS had actually gone into TIAA/CREF rather than to the federal government when Reagan and Bush could raid our funds.
          Thus SS recipients have subsidized the federal government, and now the right wing wants to steal even what is left.

          Social Security (SS) is a regressive tax – as income increases after a high middle class income, NO more tax is added. Thus if you are rich, the % you and your employer pay for SS is far lower than a middle class person pays. That is unfair. Also if you get income from capital gains, there is NO SS tax. Grossly unfair, especially considering that capital gains is taxed at half the normal rate in any case

    • I don’t see any useful conclusions to be drawn from this data. Are there things about Mississippi and West Virginia that lead to them receiving a lot of federal money and paying very little taxes other than the somewhat arbitrary classification as conservative? How much control do the poor people in those states have over the amount of federal taxes they pay and the transfers they receive? Is there any indication of a causal relationship between the idelogical classification of the states and the amount of taxes and transfers?

      I really don’t see what this is trying to tell us other than making an us vs them point that helps absolutely no one.

      • @AB this is a perfectly valid point to explore. States who implement policies to help the less fortunate (California, New York) tend to be wealthy ‘blue’ states. States who implement ‘screw the poor’ policies tend to be poor ‘red’ states.

        It would be nice to know if the conservative red states were flaming hypocrites on this matter… and if so, rub their noses in a little bit ;-)

        Altho, I do see social security being a bit of a spoiler, as I mentioned above.

        • I never questioned whether state-level variations in policy and the impact those have on poverty levels, taxation, and federal transfers is a worthwhile topic to explore. But the aggregate data tells us very little. This is way too complex to just put an arbitrary liberal/conservative label on entire states and then make an X/Y graph on net federal transfers.

          Yet that doesn’t stop Krugman from writing two posts on “moochers” and making a polemic argument rather than digging further into the data.

    • I’ll second Ed on his comment. The original chart that Krugman used had the conservative states in red and the liberal states in blue, I’m not sure why you reversed the colors when you redid the chart to include population size; but given conventional use of colors these days to signal party or political leanings, I think it could be unintentionally misleading to the casual viewer for the chart to show blue states as conservative and red states as liberal these days. Thanks.

      • I see you’ve revised the chart. colors Thanks. It’s a good chart and I appreciate the time and effort you put into making it. and revising it.

    • Bubbles? really?

    • So, Utah, Wyoming, and maybe Nebraska are the only true, self-reliant ‘conservative’ states. No surprise there; the rest are just southern Democrats who have masqueraded as Republicans over the last three decades.

    • Looks like conservative states are at the top and the bottom of the chart. I would guess that something other than conservative politics is driving it.

    • If the goal is to demonstrate that states with conservative majorities tend to be hypocritical welfare cases and states with liberal majorities are magnanimous and self-reliant, then it would make sense to look at both the magnitude and the composition of the transfer payments.

      The connection between political orientation and dependence on government transfers is an interesting one, but any serious analysis should also take tax credits – like the mortgage interest deduction, the deduction for employer sponsored health insurance premiums, etc – as well as appropriations for construction, military bases, into account before attempting to draw any firm conclusions.

    • It’s interesting to watch all the energy being expended here, because the point Krugman is making (or at least so he claims) isn’t about anyone actually being a moocher; it’s about the bad faith of conservative politicians and activists who spread the claim that people who benefit from government transfer payments are all a bunch of parasites whom the hardworking among us would do best to get rid of. Krugman is doing the mote/beam thing, and everyone is arguing about precisely what kind of wood the beam is made from.

      • And of what use is that to anyone? It’s an us vs them partisan argument, less concerned with increasing knowledge and coming up with solutions than raising one’s own status. Not the type of thing one expects when you read TIE.

        And as others have mentioned along various dimensions, this is far too nuanced a subject to draw any conclusions from the aggregate data, so to use your analogy the problem is not what type of wood it’s made of, we can’t even be certain there is a beam.

        • it’s useful to squelch conservative talking points and begin a real dialog:

          Tea Party: “Blue state welfare moochers are draining our economy!”
          Sane Party: “Actually, people in conservative red states drain more federal funds than people in liberal blue states. About 25% more. Here’s proof.”
          Tea Party: “…well, I guess it’s more complicated than just one number.”

          • Given that states aren’t sentient entities themselves, but are composed of individuals – it’s meaningless to talk about a state preaching self reliance but practicing welfare dependency.

            Unless you disaggregated the transfers to determine their composition, then actually account for the ideological commitments of the individuals receiving welfare-transfers vs age dependent transfers, then you haven’t come anywhere close to establishing that “red state conservatives” are actually singing from one tune and dancing to another.

            That may indeed be the case, but data presented in the charts at the top of the post don’t come anywhere close to supporting that conclusion.

            • Why the distinction between “welfare dependent” and “age dependent” transfers? In both cases, individuals take more out of the system than they pay in. I think we can also reasonably assume that the vast bulk of these transfers are of the “age-dependent” variety, since those constitute a much larger slice of the national budget. And have you seen the results of the National Journal survey where respondents, who appear to be confused about where entitlement spending is concentrated, strongly oppose cuts to Medicare/SS and yet believe that the “government taxes workers too much to fund programs for people who could get by without help”? Reagan’s “welfare queen” looms much larger in the imaginations of the people than is warranted by reality.

    • “You would then have the question of to what extent can people simply leave the low-wage, low-productivity places and move to the more prosperous ones”.
      First, low wage does not equate low productivity. If you don’t agree then stop accepting what our armed provide.
      Secondly, I find this statement interesting considering the EXACT opposite is happening to a large degree. With individuals and corporations. Count it to weather or right to work status but either way this attempt to, once again, paint a certain region and ideology in negative a light is transparent.
      Perhaps adding more “color” to the graph would shed more light. Maybe break down the racial make-up of the represented areas. i.e., the “low wage/productivity states have much higher pct. of African Americans. THAT wouldn’t jibe with the message we’re trying project. Better leave that out.

    • Armed = Farmers above.

    • If you change the distinction between states from liberal-conservative to Union-Confederacy, then the points for Utah, Wyoming and Nebraska move to the left, leaving a clearer upper right corner of high CTR states, all from the Old Confederacy. But what’s the underlying variable here? It could be Southern politics, but it might also be black population or some other proxy.

    • Union/Confederacy politics still predict everything, which is fairly horrifying. Why did we let those states back in the Union, again? Nothing. But. Trouble.

    • Some of the comments here appear to focus on whether these transfers are to old people or to Reagan’s “welfare queens”. Why is this question of any interest to anyone? It’s well known that the average SS/Medicare recipient takes much more out of the system than they pay in. Furthermore, the vast bulk of our entitlement spending in the U.S. goes to the elderly, not to “welfare queens”. If you’re actually driven by concerns about deficits and not by uglier sentiments, the facts compel you to be much more worried about “age dependent transfers” than “welfare transfers” (as JayB labels them). Since people of all races age, there’s no call to bring race into the discussion at all (as one poster did above).

    • Interesting – I just gave my stats undergrads an assignment using data on median age at first marriage for women and another column of divorce rates for women, had them group the data by region and draw conclusions. Many of the southern states had the highest divorce rates in the country and the lowest median age at first marriage. Many of the northeastern states had the lowest divorce rates and highest median age at first marriage. Education levels are lower in the south, wages are lower and thus poverty is higher. More poverty=more transfers.

    • Paul Krugman thinks “self-reliance” may explain Nebraska’s relative outlier status. Methinks USDA reliance may explain more. Including $14.8 billion in farm subsidies (rounds out nicely to a billion bucks a year)to a state with a population of 1.8 million would nudge the red blip only slightly north (about 1.5% back-of-the-envelope). But a little more granularity results in a much more dramatic shift.

      Take York County, for example (which I dearly love — all you need to know about America, good and bad, you can learn at Chances ‘R’ in York): 1995-2010 Subsidies $ 304 million Pop. 13,500; works out to the equivalent of personal current transfer receipts of $1500 per year per capita average. Now *that* would move the data point north about 4-5% to where Nebraska (or at least York County: 3rd Cong. Dist. Cook PVI R+24)rightfully belongs.

      (Data from Environmental Working Group’s Farm Subsidy Database http://farm.ewg.org/region.php?fips=31000)

    • By the way, I just noticed your earlier answer to Paul regarding agricultural subsidies. From what I’ve been able to discover, they are *not* included in the BEA data according to their guide (http://www.bea.gov/regional/pdf/spi2005/06%20Personal%20Current%20Transfer%20Receipts.pdf). That’s why I submitted my earlier comment. Am I incorrect?

    • Didn’t see Indiana, or any other Great Lake States on there. Did you just cherry pick your states to match your political opinion?

      • You’re kidding, right?

        Those are the ten most conservative and ten most liberal states, as ranked by Gallup. That’s how they were picked.

        Next time you can try and follow the links and read the explanations first.

    • It’d be interesting to see where the so-called ‘purple’ states fall on this chart. Do they cluster around the American average? Mix in with other, more extreme states? Form their own, distinct cluster?