Computer Chronicles Revisited 49 — Solon, Bob Carr, Ed Zschau, and F-15 Strike Eagle

Many Computer Chronicles episodes to this point have discussed, or at least mentioned, the influence of politics on the tech industry. But this next episode from November 1985 looks at how the computers were influencing politics. And our hosts were not necessarily encouraged by what they saw.

Would the “Information Age” Lead to a “1984-ish” Scenario?

Stewart Cheifet did his cold open from the United States Capitol building in Washington, DC. He said the federal government cranked out millions of pieces of paper per day. You couldn’t find a better prospect for the “paperless office” and computers. He added that computers were not only being used in government, but also in campaigning.

In the studio, Cheifet showed Gary Kildall the Washington Alert Service, an online database operated by Congressional Quarterly, Inc., which cost about $100 per hour to access. Cheifet said the service offered an incredible amount of information about what was happening in Washington. For example, you could see the exact schedules for the House and Senate or any of their committees, check on the status of a particular bill, or look at the voting records of members and their constituents.

Cheifet said this service was a good example of how computers were being used to handle the massive data management problems involved in the political process. But wasn’t there also concerns that computers could be used in a dangerous kind of “1984-ish” way to threaten the political process? Kildall replied this was a clear example how we were moving into the “information age,” i.e., where information was a valuable commodity. If you had quick-and-ready access to this information, you had a tremendous competitive advantage.

Kildall cautioned that if there was value in having this kind of background information about a politician, then you could only imagine the value of having everyone’s own personal records accessible. So how did we keep that from happening? Cheifet agreed that was a problem.

California Legislature Adopted Computerized Data Management

Wendy Woods presented her first remote segment, reporting from the California State Capitol in Sacramento, where she narrated some B-roll footage of legislative staffers using computer systems. She said that “politics as usual” was giving way to “automation as usual.” At almost every level of legislative life, computers had arrived.

In an effort to trim the 500 million pages generated by the California legislature each year, Woods said the Capitol’s data center had established a network of mainframes and personal computers to provide instant access to legislative activities. Unique bill-tracking software allowed representatives to watch the progress of legislation through the Assembly, the Senate, and their myriad of committees. The system displayed proposed amendments in the same way as a traditional paper document–but with the advantage of highlighting the amended text in color.

As members became accustomed to their new machines, Woods said, they found their own applications. In one unidentified legislator’s office, letters from constituents were coded and then matched by topic to the bill-tracking database. When a controversial bill approached the crucial vote, warning letters were quickly dispatched to those interested voters.

While some critics said a database of “choice” voters might seem like an open invitation for campaign activities, Woods said legislators were eager to point out their genuine need for the system. On average, each officeholder had 8,000 bills to watch over and several hundred thousands constituents–each one with an opinion to express.

Political Hacks Used Computer Algorithms to Target Voters Even Back in 1985

Two political campaign consultants, Roger Lee and Frank Tobe, joined Cheifet and Kildall for the episode’s first round table. Kildall asked Lee for a general overview of how computers were being used in political campaigns. Lee said the main goal was to target voters for candidates. His company, Capital Data Communications, built databases using registered voter files. It then matched that data to other files to gather additional information such as the voter’s occupation, age, ethnicity, how often they voted, and so forth.

Kildall asked Tobe if that was also what his firm–Below, Tobe & Associates–did as well. Tobe said that developing a database was just the beginning. You then gave that database to a campaign strategist so they could see the different segments that made up the candidate’s district. The strategist then spent campaign funds to try and develop these segments into people who wanted to vote for their candidate. This required the use of pollsters and other direct contract methods. Once you identified a segment, you started using different types of products to get the campaign’s message across.

Cheifet said this sounded like an “inverse if not perverse” process where you go out and find what the constituent wanted and that’s what the candidate would become. Tobe said that wasn’t exactly what happened. The goal was to find out the components that made up the candidate’s district, and then you determined how the candidate could address each of those different segments. For example, if you had a “highly ethnic” district, then it was incumbent on the candidate to have something unique to say to those ethnic groups. And that may not be the same thing that a “non-ethnic working guy” would pay attention to. So you had to identify these people just to get their attention.

(I will break the recap at this point to note that Frank Tobe was a Democratic Party political consultant. His firm took in about $5 million during the 1986 election cycle, according to the New York Times. Tobe also bragged to the Times that he had a “gay algorithm” that was “used to locate people who are likely to be homosexual, and thus interested in a politician’s stance on gay rights.”)

Kildall next asked Lee to demonstrate Solon, a software product for IBM PCs that many small campaigns used. The software had different components. One section was for incumbent officeholders and allowed them to track casework–i.e., letters or calls received by the office on a specific topic. Another section focused on campaign administration, such as the candidate’s scheduling, recording contributions, setting up polling questionnaires, performing district analyses, and so forth.

Using an IBM PC, Lee demonstrated the political office casework management segment of Solon. He said this system was currently installed in every office of the New York State Assembly. Lee pulled up a sample constituent file. The file contained the constituent’s basic contact information, whether they’d written the office before, his job, date of birth, Social Security number, ethnicity, et cetera. Cheifet asked where all this information came from. Lee said it often came from information provided by the constituent. In the case of voters, they relied on a variety of sources, such as DMV records or tax assessor lists (to see who owned property).

Kildall asked about the privacy implications of these types of databases. Lee said that all of their lists were based on publicly available information. He said the goal was to match their client’s message to the recipients. But they weren’t “prying” or using any information that wasn’t already public or otherwise secret.

Cheifet asked Tobe how important was this stuff. Was it possible that Candidate A beat Candidate B because they had better software? Tobe said there were some studies that suggested campaign tactics accounted for about 2 or 3 percentage points in an election. Most candidates won by 5 percent of the vote. So for 50 percent of campaigns, the computers didn’t have much of an effect. But for the other 50 percent it had “one heck of an effect.” So if you had products that could help you identify that “extra group of people” who will pay attention to your message, then the software could have an actual effect. Kildall noted Tobe’s firm maintained a nationwide database. Tobe said it had information on about 90 million voters, mostly Democrats.

Did Computers Lead to More Accurate Polling?

Wendy Woods returned with her second remote report, this time from Field Research Corporation, a San Francisco-based political polling firm. She said that as politicians became increasingly dependent on polls, the pollsters had become increasingly dependent on computers. For example, Field Research relied on a network of terminals connected to a Digital Equipment Corporation VAX 11/750 minicomputer. The system drew random phone numbers for operators to call and allowed interviewers to read questions right off their terminal display. The recipients’ answers were fed directly into the minicomputer, where analysts could then look at the data in a number of ways to compile their reports.

Field Research’s tabulation manager, Rick Felsing, told Woods that computers allowed them to turn things around much faster. Now when they received the data it was a matter of 2 or 3 days to generate reports. Before, it took about a week to 10 days. In fact, now they could turn around a study the moment the interviewing was complete.

Woods said that over the last few years, computers had virtually taken over the research field. But one job that computers would not be taking over soon was the role of the interviewer. Human operators still beat out the sympathized voice.

Woods added that the biggest advantage of the computer was that it produced more accurate results. Field Research’s polls, for instance, claimed just a 3.1-percent of margin of error in predicting general election outcomes.

House Members Embraced Computers While Warning About Going “Too Far”

Stewart Cheifet was back at the U.S. Capitol to interview two members of the House of Representatives who made extensive use of computers in their campaigns and congressional offices. The first subject was M. Robert (Bob) Carr, a Democratic congressman from Michigan. Cheifet noted that Carr’s office featured 4 typewriters, 7 people, and 10 computers. Carr was one of only three House members with his own computer system, which was based on IBM PCs and an Ethernet network.

Carr told Cheifet that computers had revolutionized his office. He said there were no secretaries in his office as such. All new hires were required to have some computer friendliness, and even better if they had computer training. Each staff member was therefore required to know how to type and expected to type their own letters. Nobody in his office asked someone else to do essentially their job for them.

Cheifet said Carr not only used computers to operate his office in Washington, but also to get elected back in Michigan. Carr thought that better computer systems did make a difference. He told Cheifet that he’d tried to put into his organization people who were “computer smart” and when it came to the campaign they’d run circles around the opposition.

On the other side of the aisle, Republican Rep. Ed Zschau represented a California district that included Silicon Valley. Cheifet said you’d figure that Zschau would use computers. And he did–particularly to handle mailings to voters. While some people complained about computerized letters, Zschau said it all made sense. He told Cheifet that in a campaign, the goal was to get your message to the voters. Some messages made no difference to certain voters, so you were wasting your time, energy, and money. But another message to those same voters could be very influential in their decision-making. So the use of computers to target the messages was perfectly appropriate.

Cheifet said Zschau told him that he could see the day when a computer would actually design a political campaign. Zschau said computer models of how to win a campaign based on polling and historic data could develop strategy and themes that would appeal to enough voters to get you a majority.

But both congressmen warned Cheifet about the dangers of going “too far” with computers in politics. Carr said you always had to keep the technology in its appropriate place. Computers cold help you assimilate and cross-tabulate vast amounts of data on what people were thinking. But you still could not computerize the “human element.”

Who Would Protect Voter Data from Misuse?

For the final segment, Gary Chapman joined Cheifet, Kildall, and the returning Frank Tobe. Chapman was the executive director of Computer Professionals for Social Responsibility (CPSR).

Kildall opened by asking Tobe to show an example of how he used his database system in running campaigns. Tobe showed off what appeared to be an envelope addressed by hand. In fact, it was a mass produced political mailing for an Arizona campaign. Tobe said that pollsters found that senior citizens would have the most impact on this particular campaign, which involved a ballot initiative on health care. Yet seniors were paying the least attention. So the campaigner decided to design the faux-handwritten letter and distribute 175,000 of them to senior citizens identified by Tobe’s database. Tobe said there were about 1,500 variations of this “totally personalized” letter.

Kildall asked if this particular campaign was successful. Tobe said yes. He added that about 20 different names were used to “write” the letters. These people, in turn, received about 14,000 telephone calls, Tobe claimed, even though the “senders’” numbers were never disclosed.

Kildall then turned to Chapman, whose group focused on ethical issues in computing, and asked for his feelings on someone like Tobe keeping a comprehensive database on 90 million people. Chapman said he didn’t want to “malign” Tobe or his work, but as a social question this was something that required a public debate. The idea of targeting particular constituencies to address a single issue of concern to them was “somewhat disturbing.” He thought computers had played a role in the rise of “one-issue politics,” where people were voted into office based on a single issue, which he saw as a bad trend.

Kildall asked about the sources of information, such as DMV records, which people might consider private. Chapman said there was quite a bit of confusion regarding public-versus-non-public information. Some information, such as tax records, should be public. But other information, like data in a DMV file, should not necessarily be public. What disturbed Chapman and CPSR were situations where information was gathered from a person for one purpose that they agreed to but then used for another purpose that they did not consent to.

Kildall asked Tobe how his company dealt with these issues. Tobe agreed that it was a serious problem. In his own example of a campaign targeting senior citizens, a campaign could misuse that data. He noted that DMV data was publicly available in Arizona, so they could match registered voters to their date of birth to determine if they were senior citizens. But he suggested that it would be improper to, say, target different groups of voters with contradictory messages.

Kildall added that someone could also misuse these databases to solicit contributions for a fund that didn’t even exist. Tobe said there were rules against that. For example, voter files could only be used for campaign or governmental purposes. They could not be used to sell insurance or some other commercial purpose. That’s why his own firm only handled political clients.

Kildall pushed back, noting the list still had commercial value and could be stolen. Tobe insisted they had “very serious security” to guard against theft. But Kildall pointed out, and Tobe conceded, there was no “formal mechanism” in place to protect such data.

Cheifet shifted to the subject of using computers to actually tally election results. He noted there were concerns about centralizing the tampering point for ballots. Did that concern Chapman as well? It did, he replied, adding there were currently several pending court cases concerning computer voting. These cases involved a particular computer program alleged to be especially prone to manipulation. The program used data punch cards, which could be added to a stack of legitimate votes or replaced with other cards that changed the software code used for tabulation.

Defending Yourself Against Political Manipulation

I’m just going to print George Morrow’s closing commentary verbatim, as it offered a prophetic warning:

You know, paper made it easy to record information. The printing press made it easy to duplicate information. And the computer makes it easy to manipulate information.

Politics is the art and science of manipulation–manipulating people and events. So you can expect that the computer and the politician are going to go hand and glove. Now, lobbyists have manipulated legislators for a long time, and the politicians are now going to use this computer as a device to try and manipulate you, to make you think that things are something that they aren’t.

Having a leg up on it is your best defense. And nobody every told us or promised us that a democracy was going to be easy to run. But a well-informed population, a well-informed electorate, is the best device to keep it running well. Knowing what these computers can–and will–do is your best defense.

Hackers Took Down Congressman’s Office, HBO Satellite

The Internet Archive’s record of this episode is actually a rerun from May 1986, so that’s when this “Random Access” segment by Stewart Cheifet was made.

MicroProse Profited from “Relatively Light” Civilian Casualties in 1986 Libya Bombing

There was one final “Random Access” item, which I decided to move into my discussion of the episode since there’s an interesting backstory that also plays into the overall theme of computers and politics. Once again, I’ll reprint Stewart Cheifet’s report verbatim:

Finally, a piece of game software zoomed up the charts last week, moving from 15th to 5th on the bestseller list in just one week. It’s called F-15 Strike Eagle. It’s a flight simulator that lets you pilot an F-15 attack bomber. Among the pre-programmed targets is a terrorist camp in Libya.

At first glance, and without knowing the context, this seems like a strange item. Why talk about a game that’s No. 5 on the charts? And why mention Libya?

First, let’s establish what chart we’re talking about. Billboard magazine published bestseller charts for software during this time. I looked up the chart for the week of April 26, 1986, and just as Cheifet reported, F-15 Strike Eagle had jumped from the 15th to 5th position since the prior week. In and of itself, however, that was hardly notable. In fact, the publisher of Strike Eagle, MicroProse, actually had a higher-ranked game on the chart, Silent Service, which had risen from No. 7 to No. 3. (In case you’re wondering, the No. 1 game was Origin Systems’ Ultima IV: Quest of the Avatar.)

Strike Eagle was already a well-established game by April 1986. The original release, developed by MicroProse co-founder Sid Meier for the Atari 8-bit computer line, came out in 1984. By April 1986, MicroProse had ported Strike Eagle to the Apple II, Commodore 64, and IBM PC, and the game had been on the Billboard chart for 60 weeks. So again, why did its sales ranking suddenly jump during this particular week in April? As Cheifet hinted, the answer was Libya.

More precisely, it was that two weeks earlier, on April 15, 1986, the United States armed forces conducted Operation El Dorado Canyon, a series of air strikes against targets in Libya. This operation was in turn retaliation for the bombing of a disco in West Germany on April 5 that killed three people and was blamed on the Libyan government headed by Colonel Muammar Gaddafi. At least 60 more people, almost all Libyans, died during the U.S. attack.

MicroProse–whose president and other co-founder, John “Wild Bill” Stealey, was himself a former Air Force pilot–decided to capitalize on this crisis. Now, this was long before the days of downloadable updates for computer games. So instead MicroProse issued a revised Strike Force manual in July 1986 that instructed the player on how to recreate the April 1986 “anti-terrorist airstrike” using existing scenarios from the game, including a mission based on a previous military encounter between the United States and Libya from 1981 known as the Gulf of Sidra incident. The manual update also helpfully explained how American technological superiority kept civilian casualties in the 1986 Libya bombing “relatively light.” (Civilian casualties were estimated between 15 and 30 people.)

To further Stealey’s war profiteering strategy, Wendy Woods reported for her own Newsbytes News Service on April 29, 1986, that MicroProse “spent $20,000 on a New York Times ad last Sunday, asking, ‘The World in Conflict…What Would You Do?’” Of course, MicroProse was hardly alone in marketing American militarism towards Libya during this time period. Just a couple of weeks later, on May 12, 1986, Paramount released Top Gun, a movie that featured a fictionalized air battle involving F-14 Tomcats that was likely inspired by the 1981 Gulf of Sidra incident.

Carr’s Anti-Vietnam Activism Helped Fuel Two-Decade House Tenure

American military policy also played an important role in the political careers of the two congressmen featured in this episode, Bob Carr and Ed Zschau. In Carr’s case, he was first elected to the House from Michigan’s 6th congressional district in 1974 as a staunch opponent of the Vietnam War. During his first term, Carr led an effort to end all remaining U.S. military assistance to South Vietnam and Cambodia. Carr also took a significant interest in nuclear disarmament issues, serving as an observer at the 1979 SALT II talks between President Jimmy Carter and Soviet leader Leonid Brezhnev and later making numerous inspection visits to Soviet nuclear sites.

Carr lost his third House re-election campaign in 1980 but regained the 6th district seat in 1982. During Carr’s second House tenure he served on the appropriations committee and largely focused on promoting spending on transportation projects. Carr remained in the House until 1995.

In 1994, Carr won the Democratic nomination for Michigan’s open Senate seat. Unfortunately, that was the year the Republicans gained a net total of 10 seats in the Senate. Carr lost his race to Republican Spencer Abraham, making him the only Democrat to lose a Michigan Senate election between 1978 and 2020.

Following his 1994 defeat, Carr resumed his pre-congressional career as a lawyer. In 2012, he joined George Washington University’s Graduate School of Political Management as an adjunct professor. Since 2013, he’s also been an executive in residence with Brookings Executive Education and a senior advisor with Transnational Strategy Group, LLC.

Zschau “Dodged a Bullet” with Close 1986 Senate Loss

The other congressman from this episode, California Republican Ed Zschau, also saw his House career end with an unsuccessful Senate campaign. Zschau didn’t spend as much time in politics as Carr. As you might expect from a man who once represented Silicon Valley, Zschau made his name (and fortune) in the tech industry before briefly diverting into politics.

Zschau actually started out wanting to join the Navy. After graduating from Princeton in 1961, he was accepted to the U.S. Navy Officers Candidate School in Rhode Island. But he failed the Navy’s medical exam due to a rugby injury. Zschau told podcaster Tim Feriss in 2019 that with the Navy no longer an option, he decided to get his MBA from Stanford instead. He’d go on to earn his MBA and a Ph.D at Stanford. During that time, Zschau embarked on his first career in teaching.

In 1968, Zschau founded System Industries Inc., which initially focused on providing third-party peripherals for DEC minicomputers. The company would later become known primarily for its disk drive business. Zschau served as CEO until 1981 and took the company public in 1980. System Industries remained in business until 1993, when it filed for bankruptcy.

Zschau said he was prompted to become involved in politics while serving on the board of the American Electronics Association, an industry lobbying group (and also the subject of another “Random Access” item from this episode). In 1977, Zschau led the Association’s campaign to reduce the federal capital gains tax. Those efforts bore fruit the following year when Congress cut the top capital gains tax rate from 50 to 28 percent.

Building off of that success, Zschau decided to run for Congress in 1982, winning the open 12th congressional district in California. Zschau said he never planned to serve more than three terms in Congress. He’d only end up serving two. In 1986, he decided to run for the Senate.

Although Zschau was not considered an especially high-profile congressman, he managed to spend $3.4 million and defeat a dozen rivals for the Republican nomination with just over 37 percent of the primary vote. Zschau then faced three-term incumbent Democratic incumbent Alan Cranston in the general election.

Zschau positioned himself as a libertarian-leaning Republican. But his changing views on military and foreign policy emerged as a focal point during the general election. As the Los Angeles Times reported in September 1986, Zschau pledged during his first run for Congress to oppose a costly new missile program. He then voted in favor of the program under pressure from the Reagan administration, only to switch sides again the next year.

The Times also cited Zschau’s “controversial change of heart about Arab arms sales.” Zschau served on the House Foreign Affairs subcommittee for the Middle East–a position he sought largely to champion loosening export restrictions on U.S.-made computer equipment. He’d also supported the White House’s policy of selling arms to Saudi Arabia. But after a visit to Israel, the Times said “he hesitated and finally said he wouldn’t vote for such a sale again.”

Cranston, a staunch supporter of Israel, pounced on this change of heart to paint Zschau as unprincipled. It also led to a rather unusual criminal case. One of the minor candidates on the general election ballot was Edward Vallen, the nominee of the right-wing American Independent Party. A Cranston supporter named Michael Goland, who was also pro-Israel, decided to try and divert Republican votes away from Zschau by secretly funding a series of commercials for the Vallen campaign. But the ads never disclosed Goland’s role–they purported to come directly from the Vallen campaign–and when federal prosecutors got wind, they charged him with violating campaign finance laws. After the first trial ended in a mistrial, a second jury convicted Goland.

Did Goland’s efforts make a difference? Perhaps. Zschau ended up losing to Cranston by a margin of 104,868 votes. Vallen received 109,916 votes. But Zschau apparently didn’t harbor any animosity over the loss. He later told Tim Ferriss that he’d “dodged a bullet,” as he ultimately ended up having a much greater impact on people’s lives as a venture capitalist with the firm Brentwood Associates than he would have had as a senator.

In addition to venture capital work, Zschau also had a stint as an IBM vice president in the mid-1990s, overseeing its systems storage division. During the 2000s he returned to his academic roots, serving as a visiting lecturer at Cal Tech and his alma mater of Princeton. In 2019, Zschau volunteered to serve as the interim president of Sierra Nevada College near Reno. During his tenure, which ended in August 2020, the school changed its name to Sierra Nevada University.

One final note: Zschau did briefly reappear in politics during the 1996 presidential election. Former Colorado governor Richard Lamm announced his candidacy to seek the nomination of the Reform Party that year, challenging the party’s founder, former IBM salesman H. Ross Perot. Zschau agreed to serve as Lamm’s vice-presidential running mate. Zschau later told the student newspaper at Sierra Nevada College that the campaign only lasted a couple of weeks and that it “was just a favor to Dick Lamm that I would accompany him to some television interviews.” Perot won the nomination easily in a national primary vote and selected economist Ed Choate as his running mate.

Frankly, I wouldn’t have even mentioned this footnote in Zschau’s biography if it hadn’t led to one of the stupidest coincidences in “Chronicles” history. In the last episode, I discussed how one of the guests, Nat Goldhaber, had been the 2000 vice presidential nominee of the Natural Law Party. What I neglected to mention then was that like Zschau, Goldhaber was also an unsuccessful candidate for the Reform Party’s vice-presidential nomination. During the 2000 campaign, Goldhaber and presidential running mate John Hagelin unsuccessfully tried to “double dip” and win the Reform Party nominations as well. They ended up losing after a bitterly contested convention to white Christian nationalist leader Patrick Buchanan and writer Ezola Foster.

Notes from the Random Access File