12.6/The Interpretive Lens/The Intellectual Frame

The Peace Dividend and the Hollowing, 1989–2001

The Dividend

The Bases and the Bodies

The mechanism for converting a smaller budget into a smaller institution was the Base Realignment and Closure process, a commission structure designed to take individual closures out of congressional hands. Four rounds ran in seven years. The 1988 round closed 16 major installations; 1991 announced 26 more, including Fort Benjamin Harrison in Indiana and Bergstrom Air Force Base in Texas; 1993 closed 35, among them El Toro Marine Corps Air Station in California and Lowry Air Force Base in Colorado; the 1995 round, the largest, produced 79 major closures and realignments, including the Mare Island Naval Shipyard in California with its 5,600 jobs, and the Charleston and Philadelphia naval yards.⁶ Each closure was rational at the national scale and catastrophic at the local one. Loring Air Force Base in northern Maine, listed in the 1991 round, closed in 1994; the surrounding community had built its grocery stores, car dealerships, housing stock, and school enrollments around the base, and its population fell by roughly half within a few years.⁶ California, where defense employment was most concentrated, lost military-related jobs on the order of a hundred and fifty thousand across the early-1990s rounds and suffered a recession deeper and longer than the national one⁴ — the national downturn of 1990–91 is treated elsewhere in this archive.

The people went with the installations. Active-duty strength fell from approximately 2.13 million in 1989 to approximately 1.38 million by 1999 — a reduction of some 750,000 uniformed personnel in a decade.⁵ These were not casualties. They were trained people with security clearances, technical certifications, and institutional knowledge, discharged in their thirties and forties into a civilian economy that did not always have a use for what they knew. A master sergeant who had run flight-line maintenance for twenty years, a Grumman engineer whose program was cancelled, a shipyard machinist at Mare Island: the drawdown produced hundreds of thousands of such men, and the labor market’s answer for most of them was a job below the one they had held. The dividend, at the national ledger, was real. At the individual ledger it was a severance.

Langley’s Ledger

The intelligence community took the same cut on the same logic, with one aggravating feature: what it lost was harder to count. In 1992, Congress and the Bush administration mandated an across-the-board reduction of 17.5 percent in civilian intelligence personnel, to be completed by fiscal 1997 — under three percent a year, a pace intended to be achievable through ordinary attrition. Attrition proved insufficient, and from 1993 the agencies turned to the buyout described at this chapter’s opening.¹ The reduction was therefore shaped not by any assessment of which skills the agencies would need but by which individuals chose to leave — and, eligibility being what it was, the individuals who chose to leave were the experienced.² Between 1990 and 1996 the NSA, CIA, and DIA together cut their civilian personnel by about 12 percent, somewhat short of the mandate; over the same period their current-dollar payroll rose 26 percent, because the workforce that remained was more senior and more expensive. Fewer people, higher average cost, and a distribution of expertise determined by retirement eligibility. The House intelligence committee’s own mid-decade review described a community that needed to be rebuilt even as it was still being cut.²

What the cut fell on, disproportionately, was the human side of the trade. Case officers and the language capacity that supported them — the Arabic, Farsi, and Pashto speakers whose skills took years to build — left with the buyouts, and recruiting to replace them was not prioritized; new hiring of analysts and case officers had slowed to a near halt by mid-decade. The technical side — satellites, signals interception — was maintained, and the preference was not irrational by the period’s lights. A satellite observes continuously without a handler; a human source must be recruited, paid, protected, and can produce a scandal, and the congressional oversight climate after Iran-Contra was hostile to exactly the operations that produced scandals. Technology, the decade’s dominant management faith, promised the same coverage at lower political and financial cost. From inside the period, the drawdown read as responsible stewardship of a peacetime service — trimming Cold War excess in a world whose existential adversary had dissolved. The directors who presided over it cycled quickly: Robert Gates through the first cuts, R. James Woolsey, who could not make the case for the agency’s budget in a post-Soviet Washington and resigned after two years, John Deutch, who dismissed the clandestine service he was shrinking as “a bunch of old boys,” and from 1997 George Tenet, who inherited the reduced agency and began attempting to rebuild what four years could not restore.⁷ The structural fact — visible in the period, and stated inside the community’s own reviews — was that signals intelligence is abundant and human intelligence is interpretive, and the community had chosen to grow the former while the buyouts carried the latter into retirement.²

Where the Power Went

While the institutions contracted, the power they served was used more often than at any time since Vietnam, in a geography with no organizing principle. Panama in December 1989. The Gulf in 1990–91: Iraq invaded Kuwait on August 2, 1990; the coalition of 34 nations assembled under UN authorization was the largest military alliance since the Second World War; the air campaign opened January 17, 1991, and the ground war ended in one hundred hours, with 148 American combat deaths against Iraqi military dead estimated between 20,000 and 100,000.⁸ The decision not to proceed to Baghdad left Saddam Hussein in power, a sanctions regime in place whose humanitarian cost the UN itself counted in the hundreds of thousands, and — the fact with the longest fuse — a continuing American military presence on the Arabian Peninsula. Somalia followed: 25,000 troops sent in December 1992 to protect famine relief, a mission that drifted into hunting the faction leader Mohamed Farrah Aidid, and the Battle of Mogadishu of October 3–4, 1993 — 18 American soldiers dead, two Black Hawks down, over 500 Somalis killed, the bloodiest single American firefight since Vietnam.⁹ The withdrawal came within months. Haiti, in 1994, restored an elected president by the threat of invasion and was largely forgotten. Bosnia was watched for three years while Sarajevo was shelled; at Srebrenica in July 1995, Bosnian Serb forces under Ratko Mladić executed more than 8,000 Muslim men and boys inside a UN-designated safe area, the largest mass murder in Europe since the Holocaust, before NATO’s bombing that August and the Dayton Accords in November.¹⁰ Kosovo, in 1999, was 78 days of bombing without UN authorization and without a single American combat death. The Gulf had also fixed the war’s domestic image: the first American war watched through pooled cameras and briefing-room footage in real time, a form that shaped every deployment’s coverage after it. Alongside the deployments ran the decade’s one sustained diplomatic investment — the Oslo process, opened in 1993 with a handshake on the White House lawn and closed, in effect, at Camp David in July 2000 with no agreement, while the settlement population in the territories roughly doubled across the same years.

The pattern across these actions was not strategy but oscillation: intervention when the televised images were sufficiently awful, withdrawal when the images turned. Mogadishu produced Presidential Decision Directive 25 in May 1994, which set strict conditions on future peacekeeping — articulable interests, defined exits — and the directive’s first practical consequence arrived within weeks of its signing: American non-participation in Rwanda, where approximately 800,000 people were killed in a hundred days while the administration declined to use the word genocide.¹¹

The same decade was producing the adversaries its institutions were becoming less equipped to see. Afghanistan, where the CIA had spent roughly $3 billion arming the mujahideen against the Soviets — $630 million in 1987 alone, channeled through Pakistan’s intelligence service — was dropped from American attention almost the moment the last Soviet soldier crossed the Amu Darya in February 1989; the civil war that followed destroyed more of Kabul than the Soviet occupation had, and the Taliban took Kandahar in 1994 and Kabul in September 1996.¹² The truck bomb detonated beneath the World Trade Center’s North Tower on February 26, 1993, killed six people, injured over a thousand, and had been intended to topple one tower into the other; its perpetrators — the cell around Ramzi Yousef, with the cleric Omar Abdel-Rahman later convicted of seditious conspiracy — issued explicit political demands about American support for Israel, which received almost no coverage; the attack was processed as crime, not communication.¹³ On August 23, 1996, Osama bin Laden published a declaration of war against the American presence in Saudi Arabia in a London Arabic newspaper; it was not front-page news in New York.¹⁴ The embassy bombings in Nairobi and Dar es Salaam on August 7, 1998, killed 224 people and were attributed to al-Qaeda within days; the response was cruise missiles against a training camp in Afghanistan and a pharmaceutical factory in Sudan whose weapons connection was disputed almost immediately. The attack on the USS Cole in Aden harbor on October 12, 2000, killed 17 sailors and drew no military response at all. Each attack was more lethal than the last. The agency tracking the organization behind them had established a dedicated bin Laden station in 1996, and its reporting grew steadily more alarmed through the decade’s end; the President’s Daily Brief of August 6, 2001, carried the title “Bin Ladin Determined To Strike in US.”¹⁵ What the community did not have was what the drawdown had spent: human sources inside an organization that, knowing American signals capability, had moved its operational planning to couriers and rooms.

What Filled the Space

The Cold War had been, among other things, a domestic organizing principle: a bipartisan frame that overrode internal disagreement. When it dissolved, the energy it had organized outward turned inward, and the turn had a date and a text. Pat Buchanan challenged a sitting president of his own party in the 1992 New Hampshire primary on an explicitly America-first platform and took 37 percent of the vote; six months later, on August 17, 1992, at the Republican convention in Houston, he named the new dispensation: “There is a religious war going on in this country. It is a cultural war, as critical to the kind of nation we shall be as was the Cold War itself, for this war is for the soul of America.”¹⁶ The formulation was exact — the culture war offered as the Cold War’s successor, the external enemy replaced by an internal one. The battlegrounds that absorbed the displaced energy across the following decade — the arts-funding fights, the military’s compromise on homosexual service members, the Oklahoma City bombing, the impeachment — are the subject of another chapter of this archive; what belongs here is only the budgetary observation that the nation’s organizing hostility was not retired in the 1990s. It was reallocated.

The Same Arithmetic Elsewhere

The management logic that ran the drawdown was not invented at the Pentagon, and it did not stop there. In December 1991 General Motors announced it would close 21 North American plants and eliminate 74,000 jobs over four years; the 1992 boardroom coup that followed accelerated the process, and further rounds of reductions ran on through the decade.¹⁷ The overcapacity was real; the execution destroyed labor relationships and production knowledge accumulated over decades, and the distinction between the two never appeared on a quarterly statement, because only one of them was measurable.

American journalism ran the same numbers on itself. At the height of international coverage in the 1980s, even regional papers kept foreign bureaus — the Chicago Tribune had offices in Baghdad, Buenos Aires, and Moscow — and by the late 1990s the broadcast networks had contracted to a handful of overseas bureaus each: five for ABC, three for CBS, seven for NBC.¹⁸ The newsmagazines’ retreat was measured by the Shorenstein Center: between 1985 and 1995, foreign reporting fell from 24 to 14 percent of Time, from 22 to 12 percent of Newsweek, from 20 to 14 percent of U.S. News; Time ran eleven foreign-affairs covers in 1987 and none in 1997.¹⁹ Editors explained the arithmetic without embarrassment — foreign covers were newsstand losers. The replacement model was the stringer and the parachute: why maintain a Tokyo bureau when individual stories could be bought singly? The answer — that the stringer did not carry fifteen years of source relationships, could not tell the significant from the staged, could be fooled — was true and unmeasurable, and the unmeasurable lost. The working life of the correspondents who remained, the satellite phones and the fixers and the Sarajevo rotations, is chronicled elsewhere in this archive; so are the economics of the printing crafts, which were running their own contraction in the same years. The structural point is singular: the American media’s retreat from the world coincided exactly with the decade in which American power was most actively rearranging it.

The universities had discovered the same substitution a generation earlier and perfected it in the 1990s. In 1968 roughly 20 percent of American college faculty were part-time; by the century’s end the contingent share of the profession had roughly doubled, with part-time appointments alone approaching four in ten.²⁰ An adjunct cost about a third of a tenure-track line per course, carried no benefits and no implicit promise of permanence, and could be released at semester’s end. What the tenure line had transmitted — the department’s standards, its institutional memory, the knowledge of what had been tried and why it failed — did not travel through instructors teaching one course each at three institutions. The same trade as Langley’s: the salary line saved, the formation eliminated, the saving visible and the loss not.

Magazines and newspapers ran the arithmetic on their own mastheads. A staff writer at a national monthly cost, fully loaded — salary, benefits, office, expenses — somewhere in the range of $75,000 to $90,000 a year in period dollars; a freelancer could deliver the same word count for $25,000 to $40,000, with no benefits, no office, and no cost at all in the months between assignments. The economics were compelling, and across the decade the staff positions thinned and the freelance budgets grew. What the trade surrendered was the same unmeasurable: staff writers accumulated knowledge of what the publication was and was for; freelancers read the last few issues and approximated the voice. Advertising supplied a second, quieter pressure. Ms. magazine abandoned advertising entirely in 1990 because the demands had become unsustainable — food advertisers expecting recipes on facing pages, cosmetics advertisers expecting beauty coverage in the issue, and Gloria Steinem published the accounting of it that summer.²¹ At most publications the mechanism was subtler and slower: a call from an advertising director to a publisher after a story ran, a remark passed to an editorial director, filed and unforgotten, and a decision a year later that happened to align with the advertiser’s preferences, with no causation anyone could point to. Ownership completed the frame. A family-held publication answered to an owner’s values; a publication inside a public conglomerate, or one carrying leveraged-buyout debt, answered to a quarterly earnings report — not directly, since no conglomerate executive telephoned an editor about foreign correspondents, but through the budget, which was set by people who did answer to it, and which determined everything the editors could do.

The defining property of all of it — the agency, the automaker, the newsroom, the department, the masthead — was invisibility from outside. A magazine in 1994 kept its offices and its product; what had changed was internal. The senior editor who knew why certain stories worked had retired unreplaced; the relationships had left with a departing colleague; the research desk’s budget line had quietly closed. None of it showed in the product until the product was worse, and by then no single decision could be blamed, because no single decision had done it.