It's a big election year, and one party's candidate is the successor to a popular two-term president. A little-known company offers the other party, which is in disarray, technology that uses vast amounts of data to profile voters. The election is incredibly close — and the long-shot candidate wins.

This was 1960, not 2016, and the winning ticket was John F. Kennedy, not Donald Trump.

The little-known — and now nearly entirely forgotten — company was called Simulmatics, the subject of Harvard University historian and New Yorker writer Jill Lepore's timely new book, If Then: How the Simulmatics Corporation Invented the Future.

Before Cambridge Analytica, before Facebook, before the Internet, there was Simulmatics' "People Machine," which was, in Lepore's telling:

"A computer program designed to predict and manipulate human behavior, all sorts of human behavior, from buying a dishwasher to countering an insurgency to casting a vote."

Lepore unearths Simulmatics' story and makes the argument that, amid a broader proliferation of behavioral science research across academia and government in the 1960s, the company paved the way for our 21st-century obsession with data and prediction.

Simulmatics, she argues, is "a missing link in the history of technology," the antecedent to Facebook, Google and Amazon and to algorithms that attempt to forecast who will commit crimes or get good grades. "It lurks behind the screen of every device," she writes.

If Then presents Simulmatics as both ahead of its time and, more often than not, overpromising and underdelivering. The company was the brainchild of Ed Greenfield, an advertising executive straight out of Mad Men, who believed computers could help Democrats recapture the White House.

He wanted to create a model of the voting population that could tell you how voters would respond to whatever a candidate did or said. The name Simulmatics was a portmanteau of "simulation" and "automation." As Greenfield explained it to investors, Lepore writes: "The Company proposes to engage principally in estimating probable human behavior by the use of computer technology."

The People Machine was originally built to analyze huge amounts of data ahead of the 1960 election in what Lepore describes as, at the time, "the largest political science research project in American history."

Using surveys, Simulmatics chopped voters into 480 categories, such as "Midwestern, rural, Protestant, lower income, female," compared these against four years of election returns, and started making predictions. It's the kind of analysis we take for granted in today's world of data-driven, microtargeted political campaigns, but at the time it was new and unproven.

The company ended up producing three reports for the Kennedy campaign, although it's unclear how much impact its work had. Many of its recommendations were "fairly commonplace political wisdom among his close circle of advisers," Lepore reports, like suggesting Kennedy address anti-Catholic prejudice head-on. "There's a lot of bluster and nonsense in the archival trail left behind by flimflam men," she notes.

But that didn't stop Greenfield and his colleagues from claiming credit for Kennedy's victory, sparking fears about the computer-powered manipulation of democracy. "A secretly designed robot campaign strategist nicknamed a 'people-machine' was said today to have been put to work by President-Elect John F. Kennedy's top advisers to suggest alternative methods of influencing voters," one wire service reported. Kennedy's press secretary flatly denied it, saying, "We did not use the machine. Nor were the machine studies made for us."

Greenfield, according to Lepore, "didn't believe in bad publicity." He took Simulmatics public in 1961 and began pitching the company's prediction services to media companies, advertising agencies and the government.

Soon, Simulmatics went to Vietnam, thanks to the political connections of another co-founder, Ithiel de Sola Pool, a political scientist at the Massachusetts Institute of Technology whose research included groundbreaking work on social networks. In Saigon, the U.S. Defense Department gave the company a contract to evaluate its counterinsurgency efforts to win the "hearts and minds" of the Vietnamese population. Simulmatics made a lot of money but didn't appear to produce much of any value, and the Pentagon eventually canceled the contract.

Back in the U.S., Simulmatics tried its hand at predicting race riots in Rochester, N.Y. — with dubious results — and won a contract to contribute to the landmark Kerner Commission investigating the causes of racial unrest (its report used only part of Simulmatics' work).

Lepore weaves her narrative across continents and through time with engaging, conversational prose. Her characters' personalities, families, affairs, fights and constant gossiping come alive, thanks to extensive troves of family papers and interviews with those closest to them.

At the same time, she braids in the larger context: the fracturing of the Democratic Party over the civil rights movement, the drama of the Cuban missile crisis, the upheaval of the anti-war movement on college campuses, the shattering impact of the assassinations of John F. Kennedy, Martin Luther King Jr. and Bobby Kennedy, the death of midcentury liberal idealism.

But at the heart of the book is a dissonance that Lepore never really resolves. How much didSimulmatics matter? Was it "effective but sinister," as portrayed in a bestselling thriller by Eugene Burdick, a political scientist who had worked with Greenfield? Or, as the Kennedy campaign contended, was it "ineffective and duplicitous"?

Much of the evidence points toward the latter. Simulmatics went bankrupt in 1970. "Data was scarce. Models were weak. Computers were slow," Lepore concludes. "The machine faltered, and the men who built it could not repair it."

Nevertheless, she argues, its ideas lived on:

"By the early twenty-first century, the mission of Simulmatics had become the mission of many corporations, from manufacturers to banks to predictive policing consultants. Collect data. Write code. Detect patterns. Target ads. Predict behavior. Direct action. Encourage consumption. Influence elections."

The persistent belief that human behavior can be predicted is unshakable, Lepore argues, even though "human nature does not follow laws like the law of gravity, and to believe that it does is to take an oath to a new religion."

And so we keep looking to technology to predict the future, in political polling, in academic data science programs, in Silicon Valley start-ups. Today, Lepore reports, the "predictive analytics" market is worth $4.6 billion.

Even our debates today over data privacy, over the role of technology in our lives, over the morality of using data to profile and influence voters, aren't new, Lepore illustrates.

When Greenfield was first pitching his voter simulation project — what would become the People Machine — to Democratic operatives in 1959, he sent his proposal to Newton Minow, then an adviser to Adlai Stevenson, the Democratic nominee in the two most recent elections.

Minow was alarmed at the implications, Lepore writes. He sent the proposal to another Stevenson insider, asking for advice.

"Without prejudicing your judgment," Minow wrote, "my own opinion is that such a thing (a) cannot work, (b) is immoral, (c) should be declared illegal."

Copyright 2021 NPR. To see more, visit https://www.npr.org.