"Among the weblogs, the best coverage of the Churchill controversy has been in View from a Height..." -Dave Kopel, Rocky Mountain News

"In Colorado, the Rocky Mountain Alliance of Blogs is covering the hot GOP primary between beer magnate Pete Coors and former Rep. Bob Schaffer with a great deal more insight than the Denver newspapers." -John Fund, OpinionJournal.com

"The Rocky Mountain Alliance offers the best of what the blogosphere has to offer." -David Harsanyi, Denver Post
Joshua Sharf

 notify list
to receive email when this site is updated, enter your email address:
 recent posts
Blogging 27 entries
Book Review 13 entries
Business 109 entries
China 2 entries
Colorado Politics 57 entries
Decision 2008 1 entries
Finance 8 entries
Flying 3 entries
General 86 entries
Higher Ed 28 entries
History 2 entries
History 2 entries
Israel 16 entries
Jewish 15 entries
Judicial Nomination 6 entries
Katrina 13 entries
Media Bias 7 entries
Movies 8 entries
Photoblogging 2 entries
Road Trip 5 entries
Social Investing 1 entries
Vote Fraud 7 entries
War on Terror 70 entries
my other blogs
Three-Letter Monte
Oh, That Liberal Media
Rocky Mtn. Alliance

Friends of the Alliance

other blogs
One Big Swede
American Thinker
Meryl Yourish
NRO Corner
Little Green Footballs
No Left Turns
A Constrained Vision

business blogs
Carnival of the Capitalists
Cold Springs Shops
Commodity Trader
Coyote Blog
Different River
Fast Company Blog
Financial Rounds
Freakonomics Blog
Management Craft
Trader Mike
Carnival of the Capitalists Submission

business data
Inst. Supply Mgmt.
St. Louis Fed Economic Data
Nat'l Bureau of Economic Research
Economic Calendar
Stock Charts
colorado blogs
Boker Tov, Boulder
Colorado Pols
Jeff Sherman

<-?Colorado BlogRing#->

sites, not blogs
Thinking Rock Press
 help israel
Israel Travel Ministry
Friends of the IDF
Volunteers for Israel
Magen David Adom
 1939 World's Fair
1939: The Lost World of the Fair
The New York World's Fair: 1939-1940
The Last Great Fair by Jeffrey Hart
Iconography of Hope (U.Va.)
Images From the '39 Fair
 google ads
Powered by
Movable Type 2.64

January 04, 2005


If sausage-making really were like this, we'd all starve. Fact is, there were some clarifications proposed, but nothing earth-shaking. All of three people showed up. The real action is going to be Friday, when the BRP, the Blue Ribbon Panel, gets together.

I do want to thank Common Cause, though, for bringing to public attention one nightmare of a proposition. The SoS wants to amend rule 14.7.7 to make sure that, during hand recounts, that slippery demon "voter intent" is divined. I have visions of elections officials, placing the ballot, Karnak-like to their foreheads.

"Voter intent" is relatively meaningless outside of obvious efforts to fill in circles or boxes. Voters are notorious for doing all sorts of things to their ballots, none of which registers intent. It's that trojan horse of a rule which brought us Florida 2000, Washington 2004, and Who's Next? 2006.

Posted by joshuasharf at 04:01 PM | Comments (1) | TrackBack

December 09, 2004

Election Rules Blue Ribbons

I attended the first meeting of the Secretary of State's Blue Ribbon Panel on election reform. I left both optimistic and worried. Optimistic because even the more outspoken members like Sen. Ron Tupa, 25% of whose district is routinely disenfranchised by voting Republican, seemed to take the proceedings in a more dignified, serious manner than the members of the public who showed up. Worried that these proceedings could turn into a circus if not properly managed, and there were some signs of that this time.

The Secretary addressed two sets of issues directly: what to do about registration fraud, and what to do about the timelines. A number of people were shocked to find out that there are no criminal penalties for refusing to turn in a registration, for instance.

We also got to a see a show-and-tell about two items: voting centers, which Larimer County Clerk Scott Doyle instituted in 2003, and some very fancy electronic voting machines. Both look like sound ideas.

Voting centers are basically precinct-less voting. Judges either configure the machine, or hand out a paper ballot, reflecting the ballot style in a voter's home precinct. They do cost less. But the really nice thing is that with fewer places to vote, election judges can connect to the county's voting rolls electronically. So much for tooling around the county, getting a tour of the local churches and schools while you vote early and often.

We also saw touch-screen voting machines used in Nevada this year. They print out a paper trail that the voter can see, but not touch. (This prevents someone from voting, and then using their paper receipt to collect on the $5 promised for voting a certain way.) The papers are sealed, kept by the cleark, and used for either hand-recounts or statistical spot-checks. Cool stuff.

This didn't come up, but if you combine the two, and remember that by 2006 we'll have a statewide electronic voter list, you could have the whole state on such a system. That way, when an early snowstorm traps Denver voters in Durango, election judges there could configure the machines for any precinct in the state.

On the good side, the Secretary has done her homework. She's organized the various reforms. When one agitators for a local paper-based blog tried to make the case for same-day vote fraud registration, she replied, simple, "this panel is not considering same-day registration." The panel members themselves seem to reflect the need to balance access with security.

Another fellow asked about students voting, and the Secretary was less-than-sympathetic. Paraphrasing: "we've had problems with students voting here and in their home states. The legislature specifically ruled out using student IDs as a valid form of ID for registration purposes." Letters from the college administration, however, do work, but I guess the idea is to make it more trouble than it's worth to vote twice.

Naturally, some members of the public were more interested in complaining, mostly about electronic voting, than about solving the problem. And the erroneous Alan Gilbert was there, pushing his exit-poll strategy. Gilbert showed up about an hour late, and proceeded to talk more than anyone else in the room except for demonstrators. He had more to say about Ohio than Colorado. As long as the local papers insist on giving him a megaphone, the Secretary is going to waste valuable time dealing with him. On the other hand, who knows more about stealing elections than a Marxist?

Posted by joshuasharf at 03:14 PM | Comments (0) | TrackBack

December 05, 2004

Election Rules Confab

The Secretary of State has announced a Blue Ribbon panel to examine the state's election rules, and what can be done to assure greater confidence at the polls. The first meeting is December 9.

I can't say I place a lot of faith in these sorts of panels. While I doubt that anyone is going to willfully sabotage the proceedings, I also wonder at the inclusion of politicians in the place of citizens. Ron Tupa, for instance, was one of the main movers-and-shakers behind an initiative that contemplated throwing the whole election into court, had it passed.

As I've said before, ID is easy to obtain. The ID should have an address and a picture. You show up at the polls, you vote. If you can't show up, you request an absentee, but you need a reason. There should be no same-day change of address, no same-day registration-because-I-entrusted-my-registration-to-the-homeless-guy-with-the-clipboard, no large-scale "helping" of those Alzheimer's patients at the nursing home vote what's in their interest.

Once we get a statewide voting list, things should be easier, especially in a state with few large cities close to the state borders. (It's hard, therefore, for someone to organize a bus caravan from Cheyenne to Denver, while filtering hundreds or thousands of people across the bridge from Camden to Philadelphia is relatively easy.)

I'll be in contact with the Secretary of State's office, but if there's a way to submit concerns ahead of time, I'll pass it along.

Posted by joshuasharf at 01:37 PM | Comments (0) | TrackBack

November 21, 2004

Mile High Fever Swamps

It's been noted here before that when the Denver Post discovers vote fraud, it tends to be of the kind that doesn't exist. Last week, they published an editorial calling for federal intervention to solve exactly the wrong problems. Now, the Post is on the verge of following up Florida 2000 with, Diebold 2004 ("Activists protest electronic voting"). Apparently, the presence of all of 200 people at the Coloado Capitol steps on Saturday rates a story with more words than participants it's covering.

"There's no reason that any reasonable person should have any confidence in this election," said Alan Gilbert, a political science professor at the University of Denver. "We voted on machines that were provided by a private company and were not subject to public review. This is a very, very serious thing."

The electronic voting machines, used by about one-third of Americans this year, don't leave a paper trail that can be accounted for. That could have made the difference during this year's election for Democratic candidate Sen. John Kerry, said Gilbert, who said he once taught secretary of state nominee Condoleezza Rice.

President Bush's margin of victory was roughly 3.5 million votes.

In addition to protesting the new voting technology, the group said exit-poll data was off as much as six points from the official results in swing states, which could be a signal of voter fraud.

"Exit-poll data has never been inaccurate in the U.S.," Gilbert said. "This is a new level of unfairness in American elections."

Mark Blumenthal, a nationally recognized political pollster who created www.mysterypollster.com , urged patience on his website Friday for those criticizing early exit-poll data.

"We have not yet seen any empirical evidence in the exit polls to prove the existence of vote fraud, nor any evidence that the exit-poll discrepancy can be explained by any such fraud," Blumenthal said.

In the first place, Gilbert's quotes are accepted virtually uncritically. There are no specifics given, none of the theories which have already been disproven are mentioned. The easiest way to feed a fever-swamp is exactly this sort of vague, "we can't prove it, but we know what we know" sort of accusation.

The accusation seems to rest entirely on the discrepancy between exit-poll data, and actual election results. The accusers then go on to blame electronic voting machines for the discrepancy. The logical leaps here are obvious: there's no evidence that the discrepancies were any greater in places using electronic voting machines than anywhere else; exit-poll data was extremely limited in time and scope; it's automatically the voting, and not the exit-polls that were wrong, and so on.

These and other points have been covered in greater detail elsewhere.

The Post then goes on to quote Blumenthal in the least-damning way possible to those he actually opposes quite strongly. Here's a much fuller quote:

Had I written that better, I might have said: "So to summarize, if you want to explain the exit poll discrepancy, absent further data from NEP, you can choose to believe..." My point is that there are two competing theories for the discrepancy: The first is that the exit polls were slightly biased to Kerry due to a consistent pattern of what methodologists call "differential non-response" that has been evident in exit polls to a lesser degree for a dozen years (Republicans were more likely to refuse to fill out the exit poll than Democrats). The second theory is that systematic and consistent vote fraud occurred in almost every state and using every type of voting equipment. The first hypothesis seems plausible to me; the second wildly improbable.

Also, remember that Blumenthal is a pollster. He's much more concerned with trying to figure out why the exit polls were wrong than with trying to prove or disprove vote fraud.

As for Prof. Gilbert, the Post makes no mention of his own biases in the matter. Prof. Gilbert is indeed a political science professor at DU. He's also a Marxist ("power-rivalry and capitalism versus human rights and democracy"), anti-war activist, whose interest in free-speech and democracy apparently doesn't extend to those with whom he disagrees. Take a look at this past Summer's syllabus, for instance.

Now, being an anti-war activist is not, in and of itself evidence of bias. But in the course of that activism, he's ventured into conspiracy theory before:

You donít have to be too with it to see that Iraq had little to do with fighting Al Queda," said Alan Gilbert, John Evans Professor at the Graduate School of International Studies at the University of Colorado at Denver and author of "Must Global Politics Constrain Democracy?" "It isnít too hard to figure out that the oil White House and the Lockheed Martin White House are shooting for oil, not for inspections. The way it is presented, if it is your son or daughter who is being asked to fight a war on Iraq, are you going to feel good about this thing? Thatís the reason thereís a big anti-war movement." (emphasis added)

This in a story that also wildly overstates the opposition to the Iraq War, and predicts mass outpourings that never happened. As a gauge of actual public opinion, Mr. Gilbert's compass seems a few points left of true north. Of course, Marxists frequently have a hard time admitting publicly that most people don't agree with them. It would undermine the putsch. And let's just say that Marxist commitment to accurate vote-counting is somewhat tenuous.

So what we're left with is 350 words on 200 people who can't come to grips with the fact that they lost. Their flawed statistical arguments are mentioned, their leader's Marxism is unmentioned, and the one countervailing voice is watered-down.

Let's hope that the Post isn't getting ready to follow up Florida 2000 with Diebold 2004.

Posted by joshuasharf at 06:42 AM | Comments (0) | TrackBack

November 17, 2004

Provisional Ballots Counted - Mostly

The Rocky is reporting that about 80% of provisional ballots cast in Denver will be counted, and that this "falls in line with other metro counties, which had provisional-ballot acceptance rates ranging from 76 percent to 85 percent, county clerks reported."

Meanwhile, the Post manages to turn this into reason to take to the barricades:

About 12,300 Coloradans voted in this month's election only to have their emergency ballots tossed aside.

Colorado's rejection rate for controversial "provisional" ballots was twice as high this year than in 2002.

"Voters are leaving polls with stickers saying 'I Voted,' but (a fourth) of those provisionals are being tossed away," said Mary Wickersham, who worked for the election monitoring group Fair Vote Colorado. "This difference between people who vote and those whose votes actually count runs contrary to the way Americans think their elections should work."

"Tossed aside." Yes, casually thrown away. Trashcan basketball, with county clerks pretending to be 'Melo, no doubt. If they even bothered to take them out of the envelopes. If not, they probably used them to play frisbee with their dogs.

De rigeur, the Post ignores Fair Vote Colorado's genealogy.

Try to get this straight, Ms. Wiskersham: just because a ballot is cast, doesn't mean it ought to be counted. Thank goodness this election wasn't close enough for these things to matter to people who actually have money to hire lawyers. Otherwise there'd be 24/7 street theater trying to count votes by people whose only residence here in Colorado is the hotel room the party bought for them.

When provisional ballots were used in 2002, about 12% were rejected. This year, about 24% were rejected. Part of this is because the new rules were implemented to prevent people from voting provisionally in Durango when they live in Ft. Morgan. But part of it is because election judges took the easy way out when they were stumped, handing out provisional ballots as they answer to any question.

I nearly got into a argument with the "experienced" election judge I was working with. When a couple came in, and it was apparent that they had moved into the precinct, but were still registered elsewhere in Denver, she insisted they get provos. Afterwards, I wasn't sure at all that his ballot was going to count, and I spent three phone calls to the clerk's office and about half an hour making sure that it would.

The precise reasons for each rejection have not been tallied by all election clerks, but the spike in discarded ballots raises questions about whether new laws meant to keep voters from being turned away are giving some Coloradans false confidence that their voices really are being heard, advocates said.

No, what's giving them false confidence isn't the spike in rejections, it's the advertisement beforehand. Election judges aren't supposed to be making decisions about what will count and what won't. So people are handed a ballot. That couple I mentioned before? The woman was perfectly willing to fill out a provo, even though she said to me, "oh, those are the ballots that they don't count, right?" I had to assure her that since she was registered, and I had verified her registration, that her ballot would count.

Election officials said there were several reasons for the large number of rejections, but ballots frequently were discarded because the voter was not registered in the county where he or she appeared to vote.

On Election Day, Ramon Rodriguez assumed his provisional ballot cast in Boulder would be routed to the right precinct in Denver.

Rodriguez, 19, had never voted before and told election judges he was in the wrong precinct, and even in the wrong county. Upon their instruction, he cast a provisional ballot - a vote he discovered later was not counted at all. Not even for president.

"I don't know how I can trust these people," he said after voting. "It's too late now."

Rodriguez was among 100 provisional voters in Boulder County and perhaps thousands across the state whose ballots didn't count because they voted in the wrong county. Two years ago, those votes would have counted. But another 296 provisional ballots cast in Boulder belonged to people not registered at all in Colorado. The county had almost 3,000 provisional ballots that did count.

If Mr. Rodriguez is quialified to vote, he could have read a newspaper sometime in the eight weeks leading up to the election and would have been well-informed that he needed to vote in the county he was registered in.

And throwing in people who weren't even registered to vote as victims borders on the insane. One gentleman and his wife walked in to vote. She was registered, even though she hadn't voted in a while, but he wasn't. Under the rules, if he had tried to register, and could tell me when and with what voter drive, he'd get a provo, and I'd pass that information along. I gave him three chances to tell me he'd tried to register, but he wouldn't lie, and I had to inform him that, while I'd give him a ballot, it probably wouldn't count since he wasn't a registered voter. Viewed in one light, it's a heartwarming tale of a gentleman who wouldn't lie just to be able to vote. But there's absolutely no reason to be handing out ballots to people who aren't registered.

Eight counties did not count half of their provisionals. Washington County rejected a state high of 72 percent of its provisional votes. Weld County did not count almost 57 percent. And 24 counties rejected at least a third of the votes cast provisionally.

Someone needs to get Susan Greene a textbook with a bookmark on the section describing "statistically significant." Washington County had 18 provisional ballots, and rejected 13 of them. Of the 24 counties with rejection rates of 1/3 or more, here are the total number of provisional ballots cast in 8 of them: 18, 26, 30, 25, 11, 21, 16, and 3. San Juan makes the cut because one of their three provisionals didn't make it.

"What the hell," said Jay Magness, who tried to vote at one downtown Denver precinct even though he's registered in New York City.

His vote likely didn't count.

It's nice to see citizens taking their civic responsibilities so seriously. I wonder if they bothered to ask if he had voted absentee in his home state.

New rules this year required provisional voters to fill out a lengthy - and, to some, confusing - affidavit, which, when not completed correctly, caused provisional votes not to count.

The affadavit was basically a name, address, birthdate form. If that's lengthy, God only knows how these people manage to fill out their taxes. Any additional information they could provide that would help identify them was encouraged but not required.

The Post saved the truly disturbing part for last:

Despite those rules, counties applied their own standards in some cases. Conejos County, for example, counted all the votes cast in wrong precincts, while other counties did not.

And as ballots were being accepted or rejected, it was difficult for reporters and other monitors to keep track of whether election clerks applied rules fairly or consistently. Some clerks refused to allow observers to hear the judges making final decisions on whether to accept ballots, and state law failed to require clerks to allow unobstructed observation.

This is a real problem. If county clerks can't be relied on to enforce the law properly, they need to be removed, possibly in irons. It's bad enough that this whole process has undermined public confidence in the elections process, and allowed attorneys to practice in front of mirrors for the next close election. If county clerks are making it up as they go along, they're going to deserve whatever grief they get. Even if we don't.

Did I say last? No, the last word goes to Pete Maysmith, whose role in exacerbating the controversy also goes unmentioned:

Said Pete Maysmith, executive director of Colorado Common Cause, "The process is crying out for reform."

Yes, Pete. But probably not the kind of reform you would endorse.

Posted by joshuasharf at 11:17 PM | Comments (0) | TrackBack

Red Herrings in a Red State

The Denver Post isn't done with the election. Not the results, except for one excruciatingly close state House race. No, the Post has taken a look at the voting process, and doesn't like what it sees. Sadly, it's calling for solutions to problems that don't exist, and has no ideas for those that do.

While predictions of massive problems never materialized, this year's balloting demonstrated that U.S. election laws remain a muddle of federal, state and local rules that expand voter rights in some states while disenfranchising voters in others.

Congress needs to develop a rational, consistent set of rules for federal elections that are fair across the board. A judge in Ohio shouldn't give voters one set of rules while another judge interprets the law a different way in Colorado. Or Missouri.

There they go again. First, there was no evidence of disenfranchisement. There was evidence of people not following the rules, not being able to vote. Secondly, the Constitutions specifically reserves to the states the times, places, and manner of voting. The Post is calling for federal intrusion into a process where there's no evidence that the states are a problem. It's not like interstate commerce, where unified laws are necessary to a coherent economy. You can only vote in one state, just like when you get a driver's license, it's for one state.

Colorado's new rules said that voters who went to the wrong precinct had to fill out provisional ballots, not regular ballots, and that only their votes for president counted. Rules in other states sensibly allowed votes for federal and statewide offices to be counted.

Rules in Ohio said their ballots wouldn't count at all, yet somehow, that escaped the Post's notice. The fact is, the only reason that Secretary of State Davidson permitted Presidential votes to count was that she believed she had to. She never asked for the judge to rule on this, and Judge Hoffman didn't express an opinion. The 6th District Court of Appeals ruled that Ohio didn't have to, but their ruling isn't binding here. Somehow, I doubt that if Congress handed down a law consistent with Ohio's rules that the Post would be happy with uniformity.

The lack of federal guidance was exacerbated by the success of voter-registration drives. In many states, including Colorado, an unknown number of those registrations were never turned in, and voters didn't know until Nov. 2 if they were registered properly.

Actually, to the extent there was a problem, it was the voter-registration drives' lack of success in turning in registrations. Colorado tried to get rid of paid solicitations for voter registration drives, and the Supreme Court ruled it unconstitutional. How Congress is supposed to solve this problem is beyond me. In any case, a simple call to the county Clerk's office, or if they had moved, to the clerk's office in their old county, should have sufficed.

Congress also needs to allocate the rest of the $4 billion authorized under HAVA for states to upgrade their equipment and develop the electronic voter databases mandated by federal law. So far, only about $1 billion has been allocated. Colorado is not among the 15 states that have a statewide database. Secretary of State Donetta Davidson says Colorado will have such a system in place by 2006, which should help avoid multiple registrations.

This is flatly untrue. Colorado is well ahead of the curve in getting a statewide list up and running. It's just that there's not a real-time turnaround, and we can't guarantee that the counties properly purge their roles. In any event we're not too good about purging dead people and those who have moved out of state, which seems to point to a need for a law directing the county clerks to do this more aggressively.

While an estimated 120 million voters went to the polls this year, more than any year since 1968, a huge number of people still don't vote. Government watchdog group Common Cause has raised the possibility of asking Congress to make Election Day a national holiday, possibly tied to Veterans Day, in an effort to increase voter participation. Many voters were late for work because of peak-hour congestion and long lines at the polls.

This is a red herring. The more marginal voters we register, the lower the percentage turnout will be, by definition. The reasons for voter apathy are legion, not the least of which is the sort of judicial, bureaucratic, and federal activism the Post is advocating here.

As for the "congestion," I didn't see it. There was a line at opening, but no one had to wait more than about 15 minutes. There was no line during lunch, and by closing the place was a mausaleum. Anyone who couldn't wait early had plenty of time in the evening. Over 40% of the precinct I worked at voted early or requested an absentee ballot.

The Post claims that a "good start" would be for Congress to mandate uniform rules. A better start would be for Congress to butt out, and for the state to tighten its ID requirement.

Posted by joshuasharf at 05:36 PM | Comments (0) | TrackBack

November 08, 2004

What Went Right?

So, why didn't Colorado run into vote fraud? I was certainly worried about it. The Secretary of State was worried about it. Eventually, the papers even worried about it.

So why didn't it happen?

Well, in the first place, the ID requirement tended to discourage a lot of people. We only had 9 provisional ballots, in a precinct with 524 registered voters. Also, Colorado has a history of clean elections at the administrative level, and assurances, unlike Detroit, of bipartisan poll-watching for the provisional and absentee ballots.

Also, Colorado has a law that voting can't begin unless two election judges, one from each of the major parties, are present. So there was no real way to gin up the machines ahead of time.

Combined, this means that any vote fraud would need to entail moving voters around, getting them to different precincts, and having them know addresses to use ahead of time. (I can say that in a precinct with 80% turnout, there wasn't a single case of somebody showing up and the register showing they'd already voted.)

When you do that, you need really good secrecy. I'm not saying it's impossible, especially in some of the rural areas. But it means a lot more work, and a lot greater risk of being detected. And the returns diminish, too.

We were also saved by the fact that the only really close race in the state was a state House race in District 23. With control of the House already settled, the only things riding on it are the political careers of a couple of local politicos with, um, limited legal budgets.

The facts remain, though, that the system still has holes. There are people who showed up to vote whose drivers' licenses didn't match their registration. They could have gone to vote in multiple precincts, in different counties, and it would have been almost impossible to catch them.

It also remains true that vote fraud is almost impossible to catch after the fact, since the nature of the crime precludes identifying the perpetrator.

Finally, the number of provisional ballots cast is disturbing. I'm sure we'll be seeing more of these in the future, and with early and absentee voting becoming more popular, the risk of errors will continue to grow.

Posted by joshuasharf at 11:03 PM | Comments (0) | TrackBack

A Civil War

Supreme Command

The (Mis)Behavior of Markets

The Wisdom of Crowds

Inventing Money

When Genius Failed

Blink: The Power of Thinking Without Thinking

Back in Action : An American Soldier's Story of Courage, Faith and Fortitude

How Would You Move Mt. Fuji?

Good to Great

Built to Last

Financial Fine Print

The Balanced Scorecard: Measures that Drive Performance

The Balanced Scorecard: Translating Strategy into Action

The Day the Universe Changed


The Multiple Identities of the Middle-East

The Case for Democracy

A Better War: The Unexamined Victories and Final Tragedy of America's Last Years in Vietnam

The Italians

Zakhor: Jewish History and Jewish Memory

Beyond the Verse: Talmudic Readings and Lectures

Reading Levinas/Reading Talmud