Really not worth archiving.




Me: Frank Lynch

(Current commentary)

These are my mundane daily ramblings.
For something less spontaneous, I maintain The Samuel Johnson Sound Bite Page (over 1,800 Johnson quotes), with a weekly essay springing from one of Johnson's quotations.


Archives for no purpose

My Amazon reviews



Friday, November 12, 2004:

A photo not to be missed. You'll probably have to scroll down to take it all in, but do so.
Link 1:35 PM Home

Can we please do away with the Generosity Index? I warned yesterday that you might hear conservatives taking up this flawed measure as a cudgel against liberals, and since then I've noticed it being written up by Andrew Sullivan and Michelle Malkin to suggest that "Blue" states aren't as generous as "Red" states. Because it's re-emerged as a red state blue state analysis, I'm going to bother to write something lengthy here.

The Generosity Index is created annually by the Catalogue For Philanthropy, an organization that wants to promote giving. It is in their best interest to make the wealthy feel as if they are not doing enough in terms of charity, and I suspect that this consideration has influenced the development of the measure, because it is extremely difficult for wealthier states to seem relatively generous under this flawed measure. Tweaks in the procedures would lead to a better description of the data, but would not promote giving so well.

Basically, the Generosity Index compares state rankings on adjusted gross income to state rankings on an estimate of charitable giving, and if a state ranks low on income but high on estimated donations, it's considered generous, but if the state ranks high on income but low on estimated donations, it's not. Each state get a difference in its ranks created, and then the states are ranked on these differences; the more a state moves up in "giving" compared to where it is in "having" — and compared to other states' movements — the more generous it is seen.

(All financial data in the discussions below come their spreadsheet, except where noted.)

Flaws (they bear repeating!):

  • The richest states cannot do well on this measure, and the poorest states cannot do poorly. This is because of the reliance on comparisons of rank data: to be deemed generous, your rank on giving has to be superior to your rank on earning, and when you are already ranking highly on earning it's extremely difficult to outrank yourself on giving. (States ranked lower for earning power have more room to improve their ranks on giving, and thus this difference in ranks — the basis for the Generosity Index — is easier for low earning states; this in turn makes it more difficult to wealthier states to do well on the Generosity Index.) Similarly on the low end: when you already rank low on earning, it's tough to do worse on giving. (Thanks to RDavis for making this point.)

    Who are the top ten states for adjusted gross income in 2002 (basis of the latest Generosity Index), and where did their votes go in 2004? Eight of the ten highest earning states went to Kerry, presenting an inherent bias against Blue states when Red/Blue comparisons are made for political gain (or bragging rights).

    1. Connecticut (Kerry)
    2. New Jersey (Kerry)
    3. Massachusetts (Kerry)
    4. Maryland (Kerry)
    5. New York (Kerry)
    6. California (Kerry)
    7. Virginia (Bush)
    8. Colorado (Bush)
    9. New Hampshire (Kerry)
    10. Illinois (Kerry)

    For completeness, the bottom ten states in adjusted gross income in 2002 (and who they voted for in 2004) are as follows... They all went to Bush, and any improvement in their rank on giving would lead to a conclusion that they are relatively generous:

    1. (50) Mississippi (Bush)
    2. Montana (Bush)
    3. West Virginia (Bush)
    4. Arkansas (Bush)
    5. North Dakota (Bush)
    6. New Mexico (Bush)
    7. South Dakota (Bush)
    8. Oklahoma (Bush)
    9. Louisiana (Bush)
    10. Idaho (Bush)
  • The estimate of a state's donations: to derive the ranks on a state's giving, the Generosity Index procedure extrapolates from all those in the state who itemized their charitable donations on their tax returns to all the taxpayers in the state, as if they gave just as much even though they didn't itemize their deductions. This procedure cannot be justified. In the case of South Dakota (where 15% itemized their donations), it means extrapolating to the remaining 85% — using an average of $3,746, or in Wyoming, where you'd be taking the average itemized claim of $6,356 and extrapolating from 17% to the other 83%. (You'd be saying that 83% of the taxpayers in Wyoming donated over $6,000 to charity without claiming it as an itemized deduction.)

    What ten states have the greatest positive extrapolation error, per taxpayer? (That is, multiplying the percentage of tax returns without itemized deductions by the average charitable deductions of those who did itemize...) And who did they vote for in 2004?

    1. Wyoming (Bush)
    2. Tennessee (Bush)
    3. Texas (Bush)
    4. Mississippi (Bush)
    5. Utah (Bush)
    6. Arkansas (Bush)
    7. Louisiana (Bush)
    8. South Dakota (Bush)
    9. Oklahoma (Bush)
    10. Alabama (Bush)

    And what ten states have the lowest positive benefit from this extrapolation? And who did they vote for?

    1. Rhode Island (Kerry)
    2. New Hampshire (Kerry)
    3. Maine (Kerry)
    4. Wisconsin (Kerry)
    5. Vermont (Kerry)
    6. New Jersey (Kerry)
    7. Minnesota (Kerry)
    8. Massachusetts (Kerry)
    9. Oregon (Kerry)
    10. Arizona (Bush)

    So hopefully from this you can see not only that the extrapolation is invalid, but it has a greater positive effect on "Red" states than on "Blue" states.

  • Using rank data: the procedure uses ranks three times to derive conclusions. Once to rank income, once to rank giving, and a third time to rank the disparity in ranks. Rank data has advantages and disadvantages. One of the advantages is that outliers can really distort averages, and ranks just treat an outlier as "#50 is only 1 worse than number 49" in the case of 50 states, or "#1 is only 1 better than number 2." But a disadvantage is that it treats all differences equally, and that differences in the middle (where differences tend to be small) are considered just as weighty as differences at the extremes; a rank difference of 2 is a rank difference of 2, no matter how big the raw difference which leads to it. When you layer this in three times, it just makes the data worse and worse to interpret.

    A simpler procedure would be to dispense with ranks completely (OK, maybe use them at the end to rank states overall), by creating a per capita percentage of dollars in donations to dollars in income. This would wipe away the bias Rdavis pointed out and demonstrated here in the first bullet. However, you'd still be left with the problem of what to do with the taxpayers who didn't itemize their charitable donations in their tax returns.

Last year I took this further, and checked the percentage to charity for the various states under various assumptions about the non-itemizing tax payers and what they gave. (I did three: assuming they gave nothing, assuming they gave $200, and assuming they gave $500.) I don't have the time to repeat it, but the results did show that Red states were slightly more generous than Blue states; yet the difference was small, and in general no state has a right to be proud, because the percentage going to charity is only about 2-3%. The chief conclusion shouldn't be that Red States are more generous, but that we all need to be more generous. (You can read last year's analysis here.)

If you see any commentators going on about this, will you please send them here? Thanks.
Link 11:54 AM Home

Thursday, November 11, 2004:

The Generosity Index is back, and it's still a bad measure of actual giving versus the opportunity. I wrote about this measure last year, and I pointed out a number of flaws.

  • It estimates charitable donations based on tax returns, and assumes that the entire states' taxpayers donate at the same rate as those who itemize their charitable deductions. Worst case is a state like Mississippi: only 20% itemize, so you're generalizing to another 80% and assuming that 80% gave over $4,000 to charity.
  • It uses rank data (comparing a state's rank in terms of its adjusted gross income to its rank on the assumed rate of giving), and then ranks the states in terms of this difference. Rank data treats all differences equally, whether it's a big one or a small one. They could have just done division and created a percentage instead of dealing with ranks, which are clunky.

I bring this flawed measure up again because, like I said, it's back, and it repeats the same mistakes. Oddly, they even defend the decision to generalize to an entire state's taxpayers from the itemized charitable deductions by pointing out that...

Although the proportion of itemizing taxpayers is relatively small, their charitable deductions do represent about 60 percent of the total estimated charitable contributions in the United States (The Urban Institute, 2001)

Many charitable deductions come from corporations, too, so I doubt that the remaining 40% all comes from individuals. But by extrapolating from the 30% who itemize to 100%, the measure clearly inflates total charity, because there isn't that much in charity moneys to account for.

Last year when this came out, several conservative bloggers noticed that "Red states" (those voting for Bush in 2000) seemed more generous than Blue states. You may hear the same thing this year. Make sure you read my post from last year.
Link 1:55 PM Home

In another effort to tear down the Constitution, the President wants the line-item veto restored, and has asked the Congress to be an accomplice in his plans to limit Congressional authority by asking them to come up with a way to do it within the bounds of the Constitution. The Supreme Court has already struck down a prior version, one that existed under Clinton; and while the Supreme Court made it clear that it gave the President too much power, Bush apparently disagrees.

It's a ridiculous idea, of course. First, politics involves compromise, but Bush prefers tyranny. Second, it gives Congress cover for its irresponsibility: it can pass laws full of pork, and members can brag to constituents about their effectiveness without actually getting anything accomplished. Third, is the President saying that even with the majorities he's got now he doesn't trust Congress?

A bill is a bill, and you can't take only the parts you like, any more than Americans can only pay taxes associated with the programs they agree with. Shivers.
Link 10:56 AM Home

I honestly don't care whether or not Bush-Cheney campaign manager Ken Mehlman is gay, or whether he'll be chosen to head the RNC. I suspect there's a good chance you feel the same way. What I do mind is the constant hoodwinking which the Republican party does to its members and politicians in order to further its goals, and if Mehlman is gay it's another instance. In this case it would be disguising diversity on what, no doubt, many consider an unpleasant dimension; this is, after all, what the flap was about when John Kerry described on of Vice President Cheney's daughters as a lesbian in a debate question regarding whether or not homosexuality was innate or a choice. It was never about her privacy, since Cheney had mentioned it himself, and she was already out; it was about trying to keep it from being better known among the rank and file.

On the other hand, you have the Republicans talking up the number of delegates and alternates at their convention who are black, when the RNC still lags far behind the DNC in inclusiveness.

Other hoodwinks they pull on themselves? Well, I've mentioned several times how so many Bush supporters have been found to be confused about basic facts regarding Iraq, WMDs, and September 11. The Bush administration could have cleared this up any time it wanted to, but chose to delay until we were already in Iraq before making so much as a peep. And there's also the Medicare bill's passage in Congress, and how an administration actuary was threatened with losing his job if he revealed the true cost estimates to Congress before it passed.

If you don't have faith in the value of your political agenda, don't pursue it. If you do have faith in it, pursue it forthrightly. But don't hide it under some cloak, making it impossible to discuss: that's tyrannical.
Link 10:46 AM Home

Wednesday, November 10, 2004:

A free treat. If you've been reading here a while, you know how much I love the Trashcan Sinatras. I've always found their songs melodic, with interesting melodic turns. Their new CD Weightlifting is available in an extended form that includes a DVD with them performing 5 songs, from when they were on KCRW's "Morning Becomes Eclectic" this past spring. Well, I found a video stream of those performances, plus talk. Not as easy to watch as the DVD, but it is free. So click and enjoy, then come back here and order through Amazon so I'll get a piece of change.
Link 4:41 PM Home

The Electoral College. Josh Marshall thinks it makes sense to dump it, and as I've written here before, I agree. Josh points out that it increases the relevance of the minority voters in states that lean heavily in a given direction, for instance, increasing the relevance of Bush supporters in Democratic-safe New York. But beyond that, it also increases the relevance of Democrats in New York, because their votes would never be taken for granted and each would be valuable. As an effect of that throughout the nation, voter turnout should increase, because every voter can see the direct value of their vote, rather than seeing it be diluted through the Electoral College (due to its "winner take all" awarding of electoral college votes).

I also think it would have an impact on the scheduling of primaries. I think — and I have to verify this — that primaries are scheduled so that "less safe" states voice their preference for the party nominee early; this makes them more likely to prefer the eventual nominee and vote with the party come November. But under a scenario that's not "winner take all," there would be no reason to give additional attention to "less safe" states.

If I'm right about the scheduling of primaries, this may be what causes the two parties to adopt candidates which are compromises for November; and while that long term objective is understandable, how many times have you heard voters say they're not really enthusiastic about either candidate? I think this would be less likely. Of course, we might wind up with candidates who are more polar opposites, due to greater need to appeal within the bulk of their party's voters.

UPDATE: According to this page, the Democratic primaries in Ohio, Florida, and Pennsylvania happened fairly late in the schedule. 19 states had primaries before Ohio; 29 before Florida; and 39 before Pennsylvania. Those three states received a lot of attention as battleground states because of both the number of electoral college votes they represented and their borderline status. Perhaps under the Electoral College system their primaries should have been earlier? Safe bet that Bob Graham would have done well in a Florida primary had he been around, and as that state's former governor and senator probably stood a better chance against Bush than Kerry did. (I admit some bias here: I have long thought Graham was a superior candidate for a number of reasons, and potential strength in Florida is one of them.)
Link 12:32 PM Home

Alberto Gonzales for Attorney General? In the wake of Ashcroft's resignation, the scuttlebutt is that the White House may name its counsel, Alberto Gonzalez, as his replacement. Want an idea of who this guy is? Here's two quick clues. One, when he served Bush in Texas as legal counsel and wrote memos to Bush on whether or not to commute the death sentences of capital criminals, those memos were one-sided and didn't really evaluate the issues from any perspective but why the sentence should be carried out. Secondly, Gonzales wrote memos which supported the President abandoning the Geneva Convention. The Geneva Convention supports our troops because it's a reciprocal agreement between nations, and it's also a good agreement because torture doesn't work. But Bush is a blind man. Wait, that's unfair to the blind.
Link 12:04 PM AM Home

Attention Deficit Disorder in the press. The lost munitions of al Qa Qaa are still every bit as much of an issue now as they were before the election, yet nothing has been written about it since then. (The one item I found through that Google news search leads in a way that suggests it was written before the election.) For some reason the Pentagon hasn't seen fit to add any information beyond its disastrous press conference from Friday, October 29. How many al Qa Qaa's are there, and how many of our troops are at risk over this failed peace process? Well, the good news is that major combat operations are over. We know that because the President told us, and this dust-up in Falluja you're hearing about is just the manufactured product of the biased, liberal press, trying to make the Preznit look bad.

Meanwhile, many are figure Zarqawi escaped Falluja before the assault even began. In all likelihood he probably figured he had a deadline to get out before the election, knowing that Bush didn't have the integrity to pursue victory in Falluja and in the election at the same time. This would, of course, be the second time in which Bush has apparently let him get away for political expedience: according to NBC and the Wall Street Journal, the US wanted to use his presence in Iraq (in an area not controlled by Saddam Hussein) as an argument for invading Iraq ("he harbors terrorists"), and passed on an opportunity to take him out in order to preserve that argument.

So now, apparently, he lives to fight another day, and direct more insurgents to kill more American soldiers. All those who support the troops, I just hope you voted for Kerry, because Bush has needlessly endangered our brave men and women.

CORRECTION: Under the spelling "al QaQaa," there has been mention in the press since election day. Most I looked at don't carry the issue further as it pertains to the processes in Iraq, and focus on the domestic political aspects. In many instances the press still looks at it as a manipulation by Kerry and/or Mohammed Al Baradei of the IAEA. Prove it.
Link 10:43 AM Home

Tuesday, November 9, 2004:

The good news from Falluja is that it's going better than expected. Fewer fatalities, fewer booby traps and so on. Still, based on past history, can America afford another catastrophic success?
Link 6:24 PM Home

It wasn't a mandate, not with 51%, but even if it were a mandate, there's a difference between an overall mandate and one on every single issue. Here's another thing McClellan said at yesterday's briefing, on Social Security program changes:

[T]he President remains firmly committed to it. It was something that was debated and discussed at length during the campaign. The American people spoke very clearly that they support the President's agenda, and that includes the President's views on allowing younger workers to invest a small portion of their Social Security funds in personal retirement accounts. And we will move forward on it, working closely with Congress.

What did the ballot look like in your state? Do you recall an opportunity to tell the President you supported one program or another? I didn't see that on mine. And since the country widely saw him as superior in fighting the war on terrorism, yet only elected him 51-48, there must be some areas where the people aren't so supportive of his agenda. Could this be one of them?

If he pursues all this stuff, how soon before his approval ratings continue their downward spiral? Will that mean the people no longer support his agenda, and would Scotty ever admit to it?
Link 4:00 PM Home

Can anyone explain to me why the US government wants to get involved in a religious issue? Not just an issue which the population sees as a religious issue, but one which the government sees as a religious issue? Aren't religious issues better addressed by churches than by the government? From yesterday's White House press briefing, Scott McClellan:

The President remains committed to doing everything he can to protect the sanctity of marriage. He believes very strongly that it is a sacred institution between a man and a woman.

"Sanctity," "sacred"... Look those terms up in any dictionary, and tell me they don't have religious meanings. Why does the government want to go here? Many churches are already firm in not recognizing, say, marriages between people who have been divorced. Is anyone asking a church to recognize the marriages of divorced people? Of course not. Has anyone suggested that marriages between divorced people threatens the "sanctity" of marriage, and proposed a Constitutional amendment to prevent it?

If the Administration really wants to pursue a ban against gay marriage, they have to drop the religious verbiage... Or atheists may be next, because since they don't recognize God, God knows their relationships make a mockery of marriage and threaten the American family.

Why isn't a line being drawn in the sand, that the argument has to be on some other basis? Sacred? Irrelevant to what government does.
Link 3:03 PM Home

The horror! It's The Return of the Serial Returner!! I read about this yesterday in the Washington Post, and heard it discussed this morning on CNN's "American Morning." Some retailers are working to limit their costs by tracking customers in their databases who return purchases too often for the stores' tastes, with a greater "opportunity" to retard their return behavior by denying them a return. (The Washington Post didn't use the reference "serial returner," I heard it from CNN's guest, but its use made me shiver — how often do you hear "serial" as an adjective in other contexts besides "killer"?)

What's the problem with this? Well, companies have return policies, which should be clearly posted where the customer will see it. Utilizing a database to reject some people's returns may amount to a specialized return policy for specific customers, and the customer needs to know about it. If you think about it, "serial returners" have an established relationship with companies regardless of any signs where excessive returns (as the company defines them) are implicitly accepted, as demonstrated by past store behavior.

The procedures discussed both in the Washington Post and on CNN are that customers won't hear about it until they try to return the purchase. That's too late: the store is altering the relationship, and needs to tell the customer before the purchase, so that the customer not only knows what to expect, but can consider it when they make their purchase. A customer might well prefer store A over store B because it's return policy up till this point has been more lenient, not because of the quality of the merchandise or the prices.

How to make it happen? Make sure the customer knows, before he/she leaves the store, that the return policy is now more restrictive for them; they can return whatever they've bought immediately if they wish, but once they leave the store, that day's purchases will fall under a more restrictive policy. It can be done by printing a message on their receipt, or with the cashier handing them a special slip or card, but the cashier needs to call out the message to the customer, like, "please make sure you read it before you leave the store." In this way, the return policy can be altered with no genuine impact on the customer, who is free to return their purchase right there and then, and take their trade elsewhere.

If the store is afraid of losing that customer's business, I wonder if they have thought ahead to the customer's likely reaction at point of return?
Link 8:59 AM Home

Monday, November 8, 2004:

Now misinforming more people than ever before. Fox News' ratings were up on election night, and they "clobbered the other cable news networks." While it seems their big concern now is that they may lose their underdog image — how do you complain about a liberal bias in the mainstream media when you're it? — we all need to be concerned with something that goes beyond marketing issues, and that's the ignorance of the American populace. A study released last year from PIPA (pdf file here) found that people who got their news from Fox were "significantly more likely to have misperceptions [about Iraq], while those who primarily listen to NPR or watch PBS are significantly less likely." Bob Herbert wonders if, rather than wringing their hands over the impact of moral values, Democrats wouldn't be better off by working on educating the population. But what do you do when you have an organization like Fox, clearly abusing the public trust, and a population apparently willing to seek comfort at its teat? Well, I guess you have to be persistent and shout louder, for one thing. Respectfully, of course, but be informed.
Link 9:46 AM Home

Funny misuse of the scientific method. Yesterday at NRO's "The Corner," Andrew Stuttaford wrote:

BLOWING SMOKE (CTD) [Andrew Stuttaford]

There is — or was — going to be a public debate in London on the health effects of passive smoking. Not now. Defenders of the notion that passive smoking does indeed have the noxious consequences so often claimed for it proved strangely reluctant to participate. The excuse given was, in essence, that they did not want to dignify the skeptics with a reply.


More likely that they recognize that trumpeting the supposed dangers of passive smoking is about politics not science, and simply cannot withstand critical scrutiny.

Their silence is thus prudent — and telling.

Here's what I found funny (if you didn't see it): Stuttaford presumes an understanding that concerns go beyond scientific conclusions, then layers on a "perhaps" and a "more likely," and on the basis of shakey suppositions (as if they had been verified), takes you to not one, but two thuses, that the silence is both prudent and telling. This is not logic, it's arrogance. If he had switched his "thus" with "if true" it would be completely different, but he's embraced certainty.
Link 8:49 AM Home

I have some good news, and I have some bad news... Unfortunately, this page isn't interactive, so you don't get to choose the order in which I post the news. But in keeping with my sunny disposition, let's start with the good news. The good news is that the stock market has been up since Bush was re-elected. That means that the very wealthy who own such a high proportion of the shares of stock in the country have done quite nicely since a little over half of America (a very little over half) voted for him). So if you encounter someone in a silk top hat and scarf on your way to your second job (you can't miss them, they look just like the dude on the Monopoly cards), give him a thumbs up and reassure him that he's doing well.

Now the bad news is that the dollar has hit a record low against the Euro, in reflection of US government deficits. What that means is that while Mr. Money Bags is doing well in the stock market, you are going to pay more for anything that's imported from Europe. Isn't that great? And the really cool thing is that the difference in the dollar vs. the Euro helps the price of Mr. Money Bags' stock prices!! Isn't that great? That's because, as the dollar declines in value, U.S. stocks become more attractive. So, while the deficit makes your kids pay for more of the present, it also has them putting more money into the pockets of Mr. Money Bags.

I'm tellin' ya, this is one way the rest of the world might grow to love Bush.
Link 12:03 AM Home

Sunday, November 7, 2004:

The U.S. Constitution is threatened by the way the victors are reacting to the election. Seriously. The orderly fashion in which Arlen Specter would normally ascend to chair the Senate Judiciary Committee might be altered if he doesn't swear allegiance to the President, and signal in advance that the Senate will abdicate its independent "advise and consent" role regarding judicial nominees; once that happens, not only will the Senate lose its independence, but the courts will be stacked with judges who reflect the President's points of view. And once that occurs, the independence of the legislative and judicial branches will be seriously eroded.

The delicate balance of the three separate branches is a cornerstone of America. If it goes, America won't be what you studied in school, nor will it have the systems which much of the world admired when democracies blossomed and patterned themselves after us.

But you can do something. As I wrote yesterday, you can persuade all your friends in Pennsylvania to contact Specter and express support and concerns. Please see yesterday's post, it matters. Specter is already being pressured by conservatives, and it needs to be countered.
Link 11:22 PM Home

David Brooks wrong again... In his column yesterday, Brooks labeled the idea that moral values affected the election as a "myth." In parts his reasoning is sound, and he's correct to suggest that it's a squishy, lovable term which people might artificially cling to in an exit poll. He's also right to note that evangelicals and so on were not a greater part of the voter base in 2004 than in 2000. But what he misses — and this is a really big thing to miss — is that those who cited moral values as their most important consideration preferred Bush over Kerry by 80-18. When the overall percentage is something like 51-48, you can't dismiss the importance of moral values even if you don't know what it precisely is. And as for the absence of a shift in the extent to which evangelicals or pro-lifers comprised a portion of the overall electorate, their preference for Bush may have been stronger in 2004 than in 2000 due to a difference in how the issues were framed. In 2004, for instance, Bush was able to talk more concretely about differences between him and Kerry on banning "partial birth abortions" (Kerry felt the health of the mother needed to be considered) as well as a federal "Laci Peterson" bill. So while it's true that moral values may not have had an impact on voter turnout, I think Brooks is being too facile in dismissing it as a motivator on whom to vote for.

Note: almost as many voters cited terrorism as their primary concern (19%) as moral values (22%), and the Bush-Kerry split was similar to that for those citing moral values.

UPDATE: Here's how "little effect" moral values had: if the 22% who cited moral values as their primary concern didn't vote, the 51-48 Bush victory turns into a 57-43 Kerry victory. I feel queasy about speculating about this, but is Brooks uncomfortable with the idea that moral values could drive an election?
Link 9:06 AM Home

Back to top.