Finally Some Action & Oversight On Data Mining

Democratic Senators Want Agency Data Mining Reports

Data mining is something that should alarm every citizen of the United States because it is an unproven technique of collecting bits and bytes about everyone included in some database somewhere.

Consider the databases we all know about. We go to the grocery store with a "loyalty card" that provides us with sales discounts but also allows the grocery store to track information about our patterns of sales. In order to get a loyalty card we must "register" by surrendering our address and some personal identifying information. The trouble is that the databases are not updated. The proof is that several of the loyalty cards my wife and I have used were registered several years ago at different addresses. Additionally, I have taken to not using the cards and allowing the cashiers to provide a courtesy entry rather than allowing the store to track my personal preferences by name or locale.

The same problem of accuracy exists with travel databases, even with the added security measures since the events of 9-11. Another set of databases that are notorious for being screwed up are those of the credit unions. We know those databases are screwed up because several states have investigated them for negative impact on individuals who have been victimized by bad credit reporting and the inability to clarify, amend or correct. These states found that the data collected by the various credit reporting agencies were at least 10% incorrect and possibly as much as 40% inaccurate.

If, as we are told by the Bush administration is done in the data mining process, we comb through hundreds of databases that have such accuracy and correctness errors, how much more corrupt will the data be? Like most things in the computerized world, the compounding of errors becomes exponential and geographical. Merging bad data to good only makes the good data unreliable. The fact is the best data and computer experts in the world have raised doubts about the reliability of not only the data and databases, but also the ability to collect data from multiple networked sources in accurate manner, and then process it in a meaningful way.

In fact, the DHS National Data Center has had tremendous difficulty coordinating data transfer and collection from the 16 networks used by national security agencies, military intelligence, and other government resources.

So, having the Dems start the process of asking serious questions about data mining and the unrestricted collection, compiling and amassing of personal data is a godsend. We can only hope that the Dems will be smart enough to place some serious restrictions on how data is collected, handled and used. I would advocate congress passing legislation that would preclude the collection of data from consumer databases created by commercial enterprises when a warrant or a national threat that can be documented under FISA rules exists. I would also limit the agencies of the federal government that can collect, process and amass such data, and how it can be shared with other agencies with and without warrants.

In the mean time, we might have to start restricting what data we care to share with anyone, especially the commercial enterprises in the world. When the big corporations start having difficulty collecting marketing and consumer data, their efforts will support better laws regarding data.

The government's mining of information from public- and private-sector databases for clues to terrorism and crime is widespread and federal agencies should regularly report to Congress on such activities, lawmakers said Wednesday.

"The overwhelming majority of these data-mining programs use, collect, and analyze personal information about ordinary American citizens," Senate Judiciary Committee Chairman Patrick Leahy, D-Vt., said during a hearing on balancing privacy and security. "We need look no further than the government's own terrorist watch list, which now contains the names of more than 300,000 individuals -- including infants, nuns and even members of Congress-- to understand the inefficiencies that can result from data mining and government dragnets."

Leahy said that "at least 52 different federal agencies are currently using data-mining technology," adding that there are "at least 199 different government data-mining programs operating or planned throughout the federal government." Despite its widespread use, Leahy said questions remain about how effective data mining is in preventing terrorism.

Leahy and Sen. Russ Feingold, D-Wis., touted a planned bill that would require agency reports to Congress on their data-mining activities. Feingold said he hopes the hearing will be the first step in "perhaps regulating this type of technology."

Former Rep. Bob Barr of Georgia said there never has been a comprehensive examination of who owns the data. Barr, who now heads a civil liberties coalition called Patriots to Restore Checks and Balances, said in a written statement that the Bush administration's data mining programs violate the Constitution and federal laws in several ways.

Barr recently switched his affiliation from Republican to the Libertarian Party in part because of his concerns over the GOP's actions on civil liberties and privacy issues.

"Predictive data mining is appropriate for seeking credit card fraud" and sending cops to a certain part of town, Jim Harper, director of information policy studies at the Cato Institute, said in testimony. However, "because of the near statistical impossibility of catching terrorists through data mining, and because of its high costs in investigator time, taxpayer dollars, lost privacy and threatened liberty, I conclude that data mining does not work in the area of terrorism."

Harper called for greater transparency of the practice.

Leslie Harris, executive director for the Center for Democracy and Technology, said in testimony that "technology has far outstripped existing privacy protections at the very time that legal standards for government access to data have been lowered."

Meanwhile, James Carafano, a senior research fellow at the Heritage Foundation, argued that traditional law enforcement models need to be modified. He said the threat of terrorism requires law enforcement to be preventive rather than reactive.

Carafano proposed rules to guide U.S. implementation of basic principles for fighting a long-term war in the electronic world. He said the development of technology should not justify authorizing new government powers.

He also said citizen representatives should authorize new, tamper-proof systems.