Thursday, September 28, 2006

Then They Chuckle At You...

Buried in today's Syracuse Post-Standard under this story about the election for the county Democratic party chair is coverage of Maffei's internal poll showing him within striking distance of Walsh.

First they ignore you.
The poll isn't causing much of a stir in the incumbent's camp, said Dan Gage, Walsh's spokesman.

Then they chuckle at you:
"It's not a 4 percent race," said Gage with a chuckle. "Have all the negative ads had an effect? Yes, you throw enough mud and some it it's going to stick, but it's not going to affect it that much."

He said the Walsh campaign has done polls that show very different results. It is the campaign's policy, however, not to release their internal polls.

The only reason I can imagine for them not releasing their own numbers is because they're not enough better than Maffei's.

How does it go next? Oh yes. Then they fight you. I'm waiting for Walsh's campaign to admit that their popular, household-word, "famous before he was born" 9-term "moderate Republican" incumbent boss has squandered a 29 point lead and is now in a dogfight with a virtual unknown.

And then you win.

I hope so, especially after hearing how "moderate" Republican Walsh yesterday voted against habeas corpus and for torture.

Monday, September 25, 2006

Walsh's Huge Lead Over Maffei Evaporates

A new poll has some great news: Benenson Strategy Group. 9/20-21. Registered voters. MoE 4.85% (March results)

25th Congressional District

Walsh (R) 44 (52)
Maffei (D) 40 (23)

According to these numbers, 9-term Republican incumbent Walsh has managed to lose a nearly 30 point lead to Democratic challenger Dan Maffei, and is now in a statistical dead heat, according to results of a new poll by Democratic pollster Pete Brodnitz of Benenson Strategy Group

Walsh's favorable/unfavorable ratio has gone from 56/20 to 51/36, pretty low for a long-term incumbent without a hint of scandal. Maffei's name recognition is still low; his favorable/unfavorable ratio has gone from 6/2 to 24/13.

Granted it's a commissioned poll, but even so these numbers come as a real shocker. Maffei's team and other parties have done a good job of tying Walsh to the rubber-stamp Republican congress.

"Jim Walsh has never been held accountable for his record of voting lock-step with George Bush. I'm exposing that record, and voters are angry about what he's been doing," Maffei said. "Once voters learn the truth about how he votes, his support evaporates."

"Voters are smart they already knew that George Bush and the Republican congress were hurting upstate New York, and now they are learning that Jim Walsh just goes along," Maffei continued.

This is excellent news and should increase Democratic chances of taking control of the House.

Thursday, September 21, 2006

Running from Bush, Walsh Changes Tune on Torture

Jim Walsh (R, NY-25) is clearly feeling the pressure.

According to this article in today's Syracuse Post-Standard, Walsh is one of  four senior House Republicans who are bucking Bush on detainee torture:

Walsh and his colleagues called for Bush and Republican leaders to maintain the U.S. commitment against prisoner abuse as spelled out in the Geneva Conventions.

"It's sort of an issue of conscience for me," said Walsh, R-Onondaga, noting that he differed with the Bush administration last year on issues surrounding the use of torture.

Quite a turnaround for a man who has voted with Bush 9 out of 10 times in the last 6 years.

Interestingly, neither the Post-Standard article nor the AP story identify any of the other "senior House Republicans" who are opposing the torture bill. Which suggests to me that the impetus for the letter must have come from Walsh's own office, meaning he's feeling the pressure from the Dan Maffei campaign, which has recently been highlighting Walsh's voting record.

There haven't been any polls released lately on this race, though I understand the Maffei campaign is working on one. My guess is the Walsh campaign has done a poll and is none too pleased with the results.

Update: The other four are identified in this AP story:

Bush also faced problems in the House, where GOP moderates Christopher Shays, R-Conn., Michael Castle, R-Del., Jim Leach, R-Iowa, and James Walsh, R-N.Y., publicly threw their support behind the bill opposed by the White House. The four Republicans told Majority Leader John Boehner, R-Ohio, that any House bill must maintain the dissidents' principles.


I haven't checked yet, but I will not be surprised to find out that Shays (CT-04), Castle (DE-AL) and Leach (IA-02) are in tough races.

Friday, September 15, 2006

A Softer World, E-mail Spying and Signal to Noise Ratio

Have you ever lied about your age, income or political affiliation on a survey just because you don't think anyone really needed to know that much about you?

Have you ever said "no, thanks" to a pollster because you worry that polls make people think there's nothing they can do to affect the outcome of an election?

Have you ever contributed to the failure of a top-level corporate plan that was devised with no rank-and-file input?

These thoughts all came to mind after I saw today's edition of A Softer World, one of my favorite web comics, created by Emily Horne and Joey Comeau:


text: i FREEDOM FIGHTER BOMBING love you whether you reply JIHAD or not
if my GODLESS WESTERN PRESIDENT letters are flagged
our love will outlive us in their SUBWAY SARIN GAS computers

The comic also made me wonder about whether there really is a vast NSA email database, and if so, what we could do about it. But first, a little about signal-to-noise ratio.

Once upon a time, before I started working for a small independent government watchdog, I worked for a small defense contracting firm as an electronic intelligence analyst. One of my jobs was to calculate the effectiveness of a radar system. Most people have a rough understanding of how a radar system works: a beam of electromagnetic radiation (ie, light, but at a higher frequency that we can't see) is shined at your car. Some of it bounces back at the radar detector. The speed of the car shifts the frequency of the light a little bit, and the radar receiver figures out from this shift just how fast you were going.

But while you're driving down the lane, there are lots of things going on around you that might also be reflecting the radar beam back to the detector. This creates noise, which is basically as everything else that the receiver picks up that isn't reflected directly back from your vehicle. If the ratio of the signal strength to noise strength is high enough, then the detector is reliable enough to identify transgressors.

Countermeasures

If you have no problem with the radar, then you can just drive the speed limit. But suppose you wanted to avoid having your speed detected, what could you do? One way would be to use a radar detector, which can alert to when the radar's on you. This is called a "passive countermeasure." The problem is, it might already be too late; if you see the radar then the radar sees you.

What you really want is to decrease the effectiveness of the radar by lowering its signal to noise ratio. You could start with your car. Most cars behave like mirrors to a radar beam: big blocks of metal that shine a lot of the energy directly back at the detector. If you had a lot of money you might be able to design a vehicle using difficult-to-detect composite materials like the Air Force uses, making sure your car's innards (like the radiator) are properly designed as well. This might be somewhat expensive, and your unique vehicle would soon be covered in the local media, making your attempts at stealth rather moot.

Another passive countermeasure would be to get a scrambler that mangles the radar signal and reflects it back with confusing information. The problem here is that acting alone you can't really return enough power to reliably confuse a well-designed receiver.

You could send out your own signal to try to jam the receiver; this is called an "active countermeasure." The problem here is that you don't necessarily know all the details of how the receiver works, so you'd have to put out a lot of power, and the receiver might be able to detect what you're doing (which is probably illegal).

Going it alone is clearly not the answer.

But what if you managed to get a lot of other people to, even those driving below the speed limit, to jam just a little bit, each one using a little power but in different frequencies, or scrambling a little bit, or modifying their cars just a little bit? All together, you might be able to increase the noise level enough to make the radar unreliable.

What does any of this have to do with NSA spying on our email?

This brings me back to today's edition of A Softer World. I'm no expert, but it's possible that the NSA software that scans email is looking for certain words and phrases (the signal) against the background noise of millions of "uninteresting" emails. What, I wonder, would happen if those triggers became so common that they no longer worked to identify the signal, but instead became part of the noise?

There's a danger of course if nobody goes along with you. After all, what do you think would happen to the first few people who started sending out emails with "Jihad meeting, bring your own anthrax" as the subject line. And of course, I'd rather see constitutional protections for our privacy, and at the very least, some kind of law banning warrantless domestic surveillance (what, we already have one of those?). Still, I can daydream about living in a society where the masses rise up and flood the net with emails titled "I'm mad as BUSH SUCKS hell, and I'm not going to take it TERRORISM PLANS anymore!"

Disclaimer: this message is for enterainment purposes only and is not intended to undermine or compromise any legal antiterrorism activity

Sunday, September 03, 2006

The War on Terror: We're Winning, So Why Aren't We Safer?

Prosecutions against suspected terrorists have dropped sharply in the last few years, according to a report on U.S. federal criminal terrorism prosecutions since 9/11 released today by TRAC. More than 9 out of 10 referrals received from various federal agencies like the FBI and DHS are turned down by the Department of Justice.

Not only that, but even when cases are prosecuted, sentences are light: of over 6400 cases referred since 9/11/01 that have been completed (that is, not still pending in some way),
  • Only 14 (one percent) received a substantial sentence -- 20 years or more.

  • Only 67 (5 percent) received sentences of five or more years.

  • Of the 1,329 who were sentenced, 704 received no prison time and an additional 327 received sentences ranging from one day to less than a year.

The DoJ would, and in fact does, say that the reason for the small number of convictions and few long prison sentences is becase they've successfully disrupted terrorist activity. According to a quote in the Associated Press story based on the report:
[Justice Department spokesman Bryan] Sierra said, prison sentences are "not the proper measure of the success of the department's overall counterterrorism efforts. The primary goal ... is to detect, disrupt and deter terrorist activities."

To me, that has as much a ring of truth as "we fight them over there so we don't have to fight them over here." But as Bill in Portland Maine likes to remind us nearly every day, the DHS threat level hasn't gone below Yellow (Elevated) since it was instituted March 12, 2002. If we've been so much better at disrupting terrorist activity since 9/11 that the median sentence has gone down from 41 months to 20 days, then doesn't that mean we're safer? It's all in the numbers they won't tell us...
"There are many flaws in the report," said Justice Department spokesman Bryan Sierra. "It is irresponsible to attempt to measure success in the war on terror without the necessary details about the government's strategy and tactics."

But DoJ officials have refused to give TRAC and other interested non-government parties those kinds of details, on the grounds that doing so might undermine anti-terrorism enforcement. Still, the details that are available tell a story very different from the one the DoJ would like you to buy: for example, most of the convictions obtained haven't even been on terrorism statutes: two-thirds have been on fraud charges. And half the convictions were obtained in the Eastern District of Viginia. Why?
Critics note that the heavy concentration of international terrorism referrals in Virginia East strains the principle that defendants should be brought to trial in the district and state where their crime occurred. They also argue that prosecutors favor bringing cases here because the juries in the area near the Pentagon naturally have a large proportion of active and retired military personnel and its circuit court of appeals is among the most conservation in the United States.

So what have we got? International terrorism is on the rise, but 90% of referrals never even get prosecuted by the DoJ, cases that do get lead to convictions end up with small sentences, even when tried in the most conservative courts in the country. But the threat level remains high.

This is patently absurd. Either the Bush administration has been successful in keeping terrorists in check, in which case the threat level should be lowered. Or else they've been incompetent at waging "the war on terra", in which case they have no more right to claim success in domestically than they have in their Iraq or Afghanistan wars of choice.

This September 11, there will no doubt be many official statements touting the successes of the War on Terror as well as the need to remain vigilant. And afraid. But even as Republican candidates scurry like rats from this sinking administration, we need to tar them (as well as certain Democrats) for their past support for using a horrible terrorist attack as an excuse for an increasingly undemocratic abuse of executive power. Reports like TRAC's kick another leg out from under the increasingly shaky table that is the GOP platform.