A fellow traveller on Twitter referenced “systems approaches regularly overreach and/or underachieve and therefore become discredited. But because they are the way the world actually works, we always come back to them.”, then in evidence talked about “(Eg overreach of crap 60s cybernetic computer modelling and Cybersyn, under-achievement of early Blair-era systems approaches)”. When I inquired more into this, he gave two fascinating examples – James Kahn modelling the Vietnam war (and geo-thermonuclear warfare), and “the disastrous modelling done for the NYC fire department which removed stations from poorer areas.”
Both of these are basically RAND Corporation work, early systems engineering? Or some variant of systems dynamics, operations research, or similar? But don’t seem to be notably ‘cybernetic’ if you ask me. Wanted to shared these examples – and the reminder that what works in systems/cybernetics/complexity can so easily get bound up with anything that doesn’t work – and ask more experienced colleagues if they know the theoretical basis of this work?
(An) original RAND report:
Review paper (1980): https://pubsonline.informs.org/doi/pdf/10.1287/mnsc.26.4.418
NY Times piece (1980): https://www.nytimes.com/2010/05/30/nyregion/30books.html
Joe Flood’s website: http://www.joe-flood.com/about-the-fires/
Source: Why the Bronx burned
Why the Bronx burned
It was game two of the 1977 World Series, a chilly, blustery October night in the South Bronx. The Yanks were already down 2-0 in the bottom of the first inning when ABC’s aerial camera panned a few blocks over from Yankee Stadium to give the world its first live glimpse of a real Bronx Cookout. “There it is, ladies and gentlemen,” Howard Cosell intoned. “The Bronx is burning.”
The scene quickly became a defining image of New York in the 1970s, a fitting summation of the decade perfect in every way but one: It never happened. Cosell, tapes of the game show, never said, “The Bronx is burning.”
“It’s a great quote, if it had been a real one,” says Gordon Greisman, who co-wrote and produced ESPN’s “The Bronx is Burning” mini-series based on the Jonathan Mahler book. “But we got all of this footage from Major League Baseball, including the entire broadcast of that game, and we went through all of it and it’s not there, because God knows if it was there we would have used it.”
More likely, the phrase was invented by New Yorkers — what the broadcaster should have said — and spun by credulous journalists.
But Cosell’s “Play It Again, Sam” moment is hardly the only myth that has sprung out of one of the darkest chapters of New York City history.
The South Bronx (along with Brooklyn’s Brownsville, Bushwick, and Bedford-Stuyvesant neighborhoods, and Manhattan’s Harlem and Lower East Side) was indeed burning. Seven different census tracts in The Bronx lost more than 97% of their buildings to fire and abandonment between 1970 and 1980; 44 tracts (out of 289 in the borough) lost more than 50%. “The smell is one thing I remember,” says retired Bronx firefighter Tom Henderson. “That smell of burning — it was always there, through the whole borough almost.”
But many of these fires were not — as was suggested then and is popular opinion now — caused by a rash of arsons. In fact, there’s a good chance that not even the World Series blaze was intentional. That fire was in an abandoned schoolhouse, there was no insurance policy for anyone to cash in on.
Hoodlums did not burn The Bronx. The bureaucrats did.
IN 1971, Mayor John Lindsay asked the FDNY’s chief of department, John O’Hagan, for a few million dollars in savings to help close a budget deficit. O’Hagan turned to a team of statistical whiz kids from the New York City-RAND Institute, a joint endeavor of the mayor’s office and the Santa Monica-based defense think tank famous for all but inventing the fields of game theory, systems analysis and nuclear strategy (and for devising a series of spectacular strategic failures in Vietnam).
NYC-RAND’s goal was nothing less than a new way of administering cities: use the mathematical brilliance of the computer modelers and systems analysts who had revolutionized military strategy to turn Gotham’s corrupt, insular and unresponsive bureaucracy into a streamlined, non-partisan technocracy.
For O’Hagan’s fire department, RAND built computer models that replicated when, where, and how often fires broke out in the city, and then predicted how quickly fire companies could respond to them. By showing which areas received faster and slower responses, RAND determined which companies could be closed with the least impact. In 1972, RAND recommended closing 13 companies, oddly including some of the busiest in the fire-prone South Bronx, and opening seven new ones, including units in suburban neighborhoods of Staten Island and the North Bronx.
RAND’s first mistake was assuming that response time — a mediocre measure of firefighting operations as a whole, but the only aspect that can be easily quantified — was the only factor necessary for determining where companies should be opened and closed. To calculate these theoretical response times, RAND needed to gather real ones. But their sample was so small, unrepresentative and poorly compiled that the data indicated that traffic played no role in how quickly a fire company responded.
The models themselves were also full of mistakes and omissions. One assumed that fire companies were always available to respond to fires from their firehouse — true enough on Staten Island, but a rarity in places like The Bronx, where every company in a neighborhood, sometimes in the entire borough, could be out fighting fires at the same time. Numerous corners were cut, with RAND reports routinely dismissing crucial legwork as “too laborious,” and analysts writing that data discrepancies could “be ignored for many planning purposes.”
Finally, the models fell prey to the very thing that technocracies are supposed to prevent, political manipulation. At the outset the RAND studies didn’t need to be manipulated — they provided what the politicians wanted without prompting. The models’ flaws all tended to make it appear that poor, fire-prone (and generally black and Puerto Rican) neighborhoods were actually over-served by the fire department, and recommended the cuts be focused in these politically weak areas. But as the cuts deepened, the models began recommending closings in wealthier, more politically active communities, an untenable development for the ambitious chief O’Hagan, who was well-connected in the Democratic clubs of Brooklyn and Queens and was later appointed fire commissioner.
“There was no question that where the commissioner kept his car was not a house that was going to be closed,” says RAND’s Rae Archibald, who was later hired as an assistant fire commissioner. “If the models came back saying one thing and [O’Hagan] didn’t like it, he would make you run it again and check, run it again and check.”
When the results still didn’t come back to his liking, O’Hagan’s men handled the problem. “Mostly we used [the RAND models] for the cuts, but if they came back saying to close a house in a certain neighborhood, well . . . if you try to close a firehouse down the block from where a judge lived, you couldn’t get away with it,” says retired chief Elmer Chapman, who ran the department’s Bureau of Planning and Operations Research. In those cases, continues Chapman, you could simply skip down the list of closings to a company in a poorer neighborhood. The models said there were less painful cuts to be made, “[b]ut the people in those [poorer] neighborhoods didn’t have a very big voice.”
As the city’s budget deficit ballooned, the RAND studies were used to close dozens more companies; in all, 50 fire units were shuttered or moved.
Fire inspections were cut by 70%; the fire marshal program was gutted; ancient rigs with outmoded safety features and rickety wooden ladders were pressed into service, and fire alarm boxes broke down by the score.
“I’d say a quarter to a third of the hydrants didn’t work,” says Jerry DiRazzo, who fought fires in the Bushwick section of Brooklyn. “You can see the way an area changes when they don’t repair a neighborhood. Every day I drove over the border from Queens to Brooklyn to go to work, and it was like this imaginary line was crossed. Almost like suddenly the sun wasn’t shining, like it was darker somehow . . . People would ask me, ‘How can you deal with this, seeing that every day?’ And I’d tell them, ‘I have a front row seat to the greatest show on earth.’ This was history being made, a city collapsing.”
DESPITE the models’ predictions of minimal impact, response times shot up and the number of fires that nearby companies fought as much as quadrupled. Citizens who lost their neighborhood firehouses protested. But by citing the supposed statistical infallibility of RAND and its computer models, City Hall was able to mollify the constituencies that really mattered. When the firefighters’ union filed a lawsuit to stop the closings, the department trotted out the models and convinced a U.S. District Court judge to threw the case out, and convinced New York Times editorial page to come out in favor of the closings and to credulously cite one RAND analyst who said the cuts would have no serious impact on coverage.
With fire rates already rising thanks to poverty, family dysfunction and an overcrowded, aging housing stock, the closings helped turn the fire problem into a scourge, consuming block after block of once densely populated, viable neighborhoods.
Thanks in large part to technological innovations like smoke detectors and fire-retardant building materials — O’Hagan’s own pet projects — the country at large experienced a 40% drop in fire fatalities from the mid 1960s to late 1970s. In the city O’Hagan was charged with protecting, though, fire fatality rates more than doubled.
Despite the conventional wisdom that arson was to blame, it was ordinary fires, caused by things like faulty wiring, errant cigarettes, and space heaters that drove the destruction. During the 1950s, city fire marshals attributed less than 1% of fires to arson. Until 1975, when the final round of fire cuts went into effect, that ratio never rose above 1.1%.
Where arson was a problem, it was largely the consequence of government intervention intended to mitigate the social consequences of the fires, namely no-questions-asked fire insurance for landlords in fire prone neighborhoods, and special welfare payments made to fire victims.
But even at its peak in the late 1970s, arson made up less than 7% of fires, and occurred primarily in already burned-out, abandoned buildings.
The fire cuts even helped lead to the Son of Sam shootings. In the mid-1970s, fire marshal Mike DiMarco was staking out David Berkowitz’s Bronx home after his yellow Ford Galaxy was spotted fleeing the scene of two trash fires set on City Island in the Bronx. “We had him under surveillance for months, watching his car late at night when we didn’t have any fires to run off to,” says DiMarco. But when Berkowitz moved to Brooklyn, the cut-to-the-bone fire marshal division dropped the tail, Berkowitz forgotten until he was arrested for the Son of Sam murders.
AS New York City faces its worst budget shortfall since it almost went bankrupt in 1975, some shadows of the RAND fire closings loom. The mayor’s initial budget plan calls for closing 20 fire companies by July 1, with more closings likely to come if other savings aren’t realized. The fire units up for closing will be announced this week.
Once again, the fire department is making cuts with computer models based on data of questionable validity, releasing incomplete and misleading statistics when it suits the department’s purposes, and refusing to release raw data so that their claims can be verified by anyone outside the department.
But FDNY spokesperson Frank Gribbon says this time will be different.
“The chiefs are looking at other factors as well,” as the models, he says. “”There’s a whole host of criteria, and then it’s the expertise of the chief officers who have to consider all of the facts and all of the data.””
Gribbon says the department doesn’t share the data behind the models, nor will it discuss the specifics of how the models work. “The public doesn’t understand,” Gribbon continued. “In terms of what the criteria [for closings] are, we’re not going to convince anybody by discussing, you know, the facts. We’re not going to convince anybody.”
Fire Commissioner Sal Cassano finds himself is in a difficult spot. On the one hand is an understaffed fire department going on as many calls as it ever has (building fires are down 50% from the 1970s, but the department now responds to more 200,000 medical emergencies every year). On the other hand is the man who Cassano, who was the chief of department before being promoted last year, owes his last two jobs to, a mayor intent on closing a looming budget gap.
Like the 1970s, firehouses are being closed while futuristic technology projects, outside consultants and computer models are still being funded. Last year the department paid computer consultants from Hewlett-Packard $3.5 million, about as much as it costs to keep two firehouses open and fully staffed for a year, to continue fine-tuning the Automatic Vehicle Locator (AVL) system they’d already installed. AVL is part of a new dispatch modeling system built by Deccan International (the same company that built the computer models being used to close fire companies), which in turn is part of a $2 billion overhaul of the city’s emergency dispatch system.
That the department needs to maintain a modern communication and dispatch system is clear, but the usefulness of spending millions to update street-corner fire alarm boxes that the department is planning on shutting down anyway, and equipping 911 operators with special software programs to receive live video feeds from callers, is questionable when basic city services are being slashed.
In a move strangely reminiscent of Rudy Giuliani’s ill fated decision to put all of his Office of Emergency Management eggs in a 7 World Trade Center-housed basket, the department is spending more than $300 million consolidate each borough’s fire dispatch office into one unit at the department’s Metrotech headquarters, and hundreds of millions more to build backup dispatch unit in The Bronx in case the Metrotech unit breaks down or is attacked.
The city has spent more than $20 million on a new Unified Call Taker (UCT) system that lets 911 call takers write down fire information and send it directly to fire dispatchers, instead of simply passing the caller along to more experienced fire call takers. Firefighters have taken to calling UCT the “U Can’t Tell” system after being sent to a series of incorrect addresses by the 911 call takers. And fire call takers are now playing a larger role in the call taking process — eliminating much of the reason for building the UCT system in the first place — after 911 call takers sent fire crews to the wrong addresses for fires in Brooklyn and Queens last November, and three people died in each fire.
A month after the fatal fires, Deputy Mayor Skyler praised UCT in testimony before the city council, saying that it “lowers response times in an effort to save lives.” But according to fire union critics, those lower response times are only true on paper, not in reality. Unlike most fire departments, the FDNY does not count the time a caller spends on the phone with a 911 operator in its response time calculations. And now that 911 operators are taking down fire information, that time is more than a minute, according the Uniformed Firefighters Association. This means that while the FDNY is reporting faster response times, the amount of time it actually takes fire crews to arrive might actually be longer.
THERE is little in New York’s current criminal, economic or building fire trends to indicate that the city will be returning to the ashen anarchy of the 1970s any time soon, but some of the management lessons to be learned from that era are clear: Whiz Kid consultants with plans to save the city through technology have their place, but shouldn’t come at the cost of basic services.
And while numbers can sometimes cut through the fog of government decision-making, they can just as easily be mistaken or manipulated.
“The models might be able to help you a little bit with closing fire companies,” says former fire commissioner Thomas von Essen, who led the department through the terrorist attacks of 9/11. “But there are so many other parts to those decisions, not just response time but the effectiveness of the unit, the political response from the neighborhood, what kind of buildings are nearby, whether there are schools or hospitals or terrorist targets.
“There’s no question that there are neighborhoods where if the firehouse is removed, it will have a minor impact. But there are also many communities that need additional fire units. It should be an ongoing process, not just something to scare the public in a fiscal crisis.”
Joe Flood is the author of “The Fires: How a Computer Formula, Big Ideas, and the Best of Intentions Burned Down New York City — and Determined the Future of Cities” (Riverhead), in stores on May 27.
Source: Why the Bronx burned