By the time I finished writing my blog post on the tragedy in Parkland, Florida on Thursday evening, I was not yet aware of the monumental FBI screw up in this case. When that news spilled forth the next day, I was distressed but not surprised. This is not the first time the FBI has ignored warnings from the public that could have saved dozens of lives. So in addition to the other non-gun control problems we face in coming to grips with the mentally ill perpetrating gun violence in our society, we can add “fixing the FBI” to the list.

Some people have used the FBI role in the tragedy to make partisan political points. Ann Coulter, for example, has said that if the FBI had not been obsessed with trying to bring down the Trump presidency, it might have spared some time to take warnings about Nikolas Cruz seriously. Of course, this is an obscene remark of the sort we have become used to from her. But she isn’t the only one to misframe the event. Florida Governor Rick Scott labeled the behavior of the FBI in this case “unacceptable,” and called on Director Christopher Wray to resign. (With the Nunes memo business, it’s amazing that the Director hasn’t resigned already anyway, as he threatened to do if his Deputy, Brian McCabe, was forced out; but that’s another story.) What we really have here with the Governor, and others, is a slightly more civilized form of political posturing. Governor Scott gives no evidence of understanding what the underlying problem is. He seems to think it’s a simple matter of outing a failed employee or two. Alas, it’s not so simple.

The basic problem is that large organizations, like the FBI, that deal with the public have a serious signal-to-noise ratio challenge. These organizations are limited in size and budget, even if both are not small. The frequency and nature of communications from the public—and that includes tips about dangerous characters like Nikolas Cruz—are such that large organizations create protocols based on probabilistic assessments that enable them to grapple with the overwhelming volume they would otherwise have to contend with. Some organizations do this efficiently and effectively, others not so much. The FBI, regrettably, seems to fall into the latter category (as does the local sheriff’s office in Parkland, Florida).

Perhaps the key issue for large organizations attempting to filter communications from the public concerns how to nurture a sense of ownership and responsibility for even small judgments. The basic rule is that the more people and the more layers a bureaucracy has, the less likely those on lower levels are to take responsibility for outcomes. No one wants to be accused of hysteria, of taking every possible threat so seriously that acting on all of them would paralyze the organization. So they tend to pass the buck, trying to make the least amount of noise while so doing, hoping that someone higher up the bureaucratic food chain takes responsibility for outcomes.

Moreover, the employees at the FBI who interface directly with the public are relatively low-status employees compared to agents and managers. Their training usually involves getting them to recognize key words or phrases in what they are hearing; they are encouraged, in essence, to count or check boxes rather than to think. Again, they are trained to listen for certain key words, and these days that means that someone out in the public who wants to get their attention had better use the word terrorism, or bomb, or Muslim somewhere in their story. That sort of standard operating procedure tends to efface a sense of agency, with the result that the key interface between the public and the FBI exhibits structural risk aversion to everything that is not a top priority.

Another way to put this is that lower-level interlocutors with the public will tend to size down threat levels to accord with their acquired perception of  institutional capabilities and priorities. The higher ups want to know about terrorism threats, about violent organized crime, lately about massive election hacking by foreign agents using U.S.-based technologies, and a few obvious others. Disturbed teenagers able to buy semi-automatic weapons make the list, but make it too often and so rank not that high on it. A good employee, one likely to be promoted and compensated more generously, is an employee who responds accordingly to priorities set above them. Someone who is always crying wolf, so to speak, or goes off on a tangent to the key priorities, will not thrive in organizations like the FBI.

Finally on this point, after 9/11 a debate over the proper siting of counterterrorist efforts in the U.S. government took place. Some, like General William Odom, argued that the FBI was very unlikely to do the counterterrorist mission well. As a part of the Justice Department, its entire culture from the very start has been directed toward apprehending criminals and jailing them. They kick into action after a crime has been committed and they surveil suspects accordingly. Preventing terrorism requires an entirely different mindset:  no crime has been committed and the constraints on surveillance and other detection methods “fit” differently from the common criminal enterprise. The conclusion: The U.S. government needs a new organization, along the lines of the British MI-5, to do domestic counterterrorism.

It was an excellent analysis and argument, but it demanded bold leadership and change—so of course it failed. The decision was to saddle the FBI with this new mission. The result was that the FBI got more budget, but not enough to cover the new missions and the old ones together. Over time the new missions essentially pushed out the status and some of the resources available for working on the older missions—missions that include preventing the kind of thing that happened in Parkland on Wednesday. Whether the FBI has yet overcome its subcultural history to do a good job at counterterrorism is a matter of opinion.

What is true for the FBI is also true for many other organizations, such as the State of Florida Department of Children and Families that judged Nikolas Cruz to be a low-risk problem despite reams of evidence suggesting otherwise. Some of this simply has to do with caseload. When an employee from such a department scores a potential problem as high risk, then a great deal of work has to be done, and a number of difficult legal and other hurdles overcome to deal adequately with that risk. All of this costs money and eats up time. So again, the institutional bias baked into bureaucratic reality is to underplay potential dangers.

Some of the legal hurdles seem frequently to reside in certain judges who refuse to take this sort of thing seriously, even though they are fast to sign no-knock search warrants for all kinds of fairly pedestrian drug misdemeanors where no violence is likely. They want to curry favor with the police, but public health officials seem only to cause them trouble. It would be useful for some intrepid researcher to examine systematically the incentive structure of local judges in order to explain the observed tendency for them to make life difficult for those who would get potential mentally disturbed, if not necessarily clinically ill, murderers off the street.

It also goes without saying that tragedies averted rarely make news, and no one can ever be certain that what an organization does to prevent a tragedy is really what is responsible for a non-event—if anyone ever asks a question in the first place about a non-event, which is even rarer. In other words, also baked into the situation is a tendency to think that a massive catastrophe is so unlikely that this particular case can’t be the cause of one, and if I fix it to make sure, the employee may well think, no one will ever know, precisely because I will have fixed it; and I will never get any credit for it.

That clarifies the imbalance in the typical incentive structure of low- to mid-level employees in such an organization: Do something and neither you nor anyone else will ever know for sure if it was the right thing to do; don’t do something and nothing is likely to happen, and you avoid scrutiny for wasting departmental resources and causing bureaucratically superior others to do more work. And if, God forbid, something does go really wrong, the likelihood is that it won’t be traceable to your decision, and even if it is, somebody higher up will get blamed if anyone gets blamed. In sum, there isn’t much of a positive feedback loop for being vigilant, and there is at least a modest benefit in a hierarchical bureaucratic setting to letting gray-area cases slide.

That does not mean that we should forgive those who make critical errors. Even as their bureaucratic environment militates against the mass of employees taking responsibility for outcomes, it is nevertheless their job not to miss these sorts of dangers. The low-ranking FBI employees who failed to transmit at least two clear and timely warnings about Nikolas Crus to the FBI’s Florida field office screwed up and should be held personally responsible for so doing.

That said, real culpability lies with the more senior directors and the ways they train and manage the personnel beneath them. Clearly, many such  directors and managers at the FBI and elsewhere are doing a poor job of instilling a sense of agency and responsibility in their mid- and lower-level employees who are first-contacts with the public. Either these senior personnel do not understand the social-psychological dynamics at play, as laid out in summary form above, or they do not know how to mitigate them. It is possible to mitigate them; lots of comparable organizations in other countries do a much better job, and it’s not rocket science t0 figure out how.

If this were a serious society, a real outcomes-based meritocracy, those more senior people who are derelict or inefficacious in their duty would be identified and either demoted, punished, or fired. And we would look to best-practice examples to improve our capabilities. But we no longer live in a serious meritocracy insofar as public service functions are concerned, so I doubt that anyone at the FBI or at the Department of Children and Families in the State of Florida will be fired for their disastrous failure to properly do their jobs in the Parkland case. No one will seriously call into question the frail existing protocols that allow such disasters to repeatedly take place. The result is that the present outrage will be wasted, things will calm down after a while, and things will go on as they have. We will need good “luck” to prevent the next massacre and hence never know how lucky we have been; typical “luck” of the sort we know only too well will go on, and on and on.

I have made the case that the FBI confronts a structural problem typical of organizations of its kind that interact with the general public, but I have also judged that, in my view, the FBI is not the most efficient or effective in dealing with the structural problems it faces. Why do I say that?

I claim no particular deep knowledge or experience in dealing with the FBI. I’ve talked to several people who do have such knowledge and experience, but beyond that I have but anecdotal experience—and as every social scientist knows by heart, anecdotes ain’t data. But since this is a blog post and not a formal essay, I will share now what little direct experience I do have with the FBI—for whatever it might be worth.

Back in 1999 and 2000 I was a staff member for the U.S. Commission on National Security/21^st Century, also known as the Hart-Rudman Commission. The commission proceeded in three parts over more than two years’ work: first to assess what the national security environment would look like twenty years into the future; second, to devise a strategy suitable for dealing with that environment; and then third, to ask whether the bureaucratic structure inherited from the Cold War was the best structure for implementing the strategy. The obvious answer was no, which is the reason the commission was created in the first place.

This third rubber-meets-road part of the commission was focused largely but not entirely on deficiencies in homeland security. The three key border agencies of the Federal government were located in three different Executive Branch departments: the Customs Agency in the Department of the Treasury; the Border Patrol in the Department of Justice; and the Coast Guard in the Department of Transportation. We had Customs and Border Patrol agents in the same districts along the Mexican border who did not even have interoperable walkie-talkies. Meanwhile, the Immigration and Naturalization Service had been tasked by Congress to do two incompatible things: let some people in, and keep other people out—with both tasks massively underfunded by the Congress.

The Commission took testimony from about three dozen Federal agencies and departments in its research phase. Of the twelve commissioners, six Republicans and six Democrats, a number would show up for every hearing to ask questions. Members of the research staff were invited to attend, so long as there they merely observed the proceedings from the back benches—after having briefed the commissioners on the situation at hand within each given agency or department, and suggested relevant questions to pose to witnesses.

I served as the chief writer for the commission, so I brought no particular specialty to the table for the organizational analysis we undertook. But I attended quite a few of these hearings over a period of a year or so, and I learned a great deal from them about how the U.S. government actually works. I distinctly remember the two fairly senior officials who came to testify before the commission from the FBI being without question the stupidest officials we hosted, and the impression was not just mine alone.

Basically, the FBI came with reams of statistics laid out neatly on charts, and they proved simply unable to imagine anything happening that had not already happened, something they had no numbers for. When first one and then a second commissioner asked them about the possibility of mass-casualty terrorism in the United States, they found it difficult to answer anything at all. They looked at their charts and numbers and shrugged their shoulders. When all was said and done, their answer was, in essence, that the probability of such a thing happening was completely unknowable because it had not already happened.

Toward the end of the session, I remember one of the commissioners present that day, James Schlesinger, looking over his shoulder at me sitting back against the wall about 15 yards away, and fixing me with a dyspeptic sort of grin that could have bent steel. Schlesinger had been CIA Director back in the day, and his impression of the FBI had been formed long before the Hart-Rudman Commission was impaneled. He had warned me what was likely to happen before we all took our seats that morning, and his grin’s message was clear: “See, what’d I tell you?”

My second anecdotal experience is much more recent. Earlier this month I learned that I was a victim of attempted identity theft fraud. Someone out there has my Social Security number, home address, birthdate, and is using my name for various illegal purposes. This person tried to steal our IRS refund this past April, tried more recently to file for my Social Security retirement benefits, and using my credit tried twice to buy a Land Rover—once in Orlando, Florida and once in a suburb of Atlanta. I learned about this because careful salesman in both car dealerships warned me what was going on.

I filed a police report with my local Montgomery County, Maryland police. I had previously filled out the proper protective form for the IRS. I was also counseled to fill out an identity theft report with the Federal Trade Commission at identitytheft.gov. Thanks to information I got from one of the car dealers, I also contacted both the Macomb, Illinois police department and the Illinois State Police district for that town—this because of a mailing address the culprit instructed the car dealer to send certain papers to for signature. From that dealership I also got photocopies of a forged Maryland driver’s license in my name linked to my home address, and a copy of the fraudster’s Geico insurance card, made out in my name and connected to my address, as well.

Of course I thought about contacting the FBI, too. So one day when I was downtown, not far from one of the main FBI buildings on 6^th Street, N.W., I decided to stop in to see if I could get any help. I was informed that the door that the FBI does not do walk-ins. This did not surprise me. Anyone with a little imagination can envisage what a zoo the building would be if just anyone could wander in with some crazy story to tell. So a guard at the door, dressed in uniform with a hat and badge and holstered gun, gave me a card to help me convey my information to the FBI. You can see what the card looks like below (the reverse side is blank).

I was cheered by the prospect of emailing my information directly to the Washington field office of the FBI. But then I noticed that the email address given for that purpose misspelled the word “field.” So I filed my report using two email addresses: one with the word “field” misspelled, and one with it spelled correctly. It didn’t matter: both emails bounced.

So I called the telephone number on the card, and quickly learned that the FBI does not accept emails. I then asked the FBI woman on the phone why, then, did someone in uniform at the FBI building on Sixth Street give me a card with an email address on it? Here is how the conversation more or less went, as best I can recall it after a two-week period.

FBI: Good afternoon, FBI.

AMG: Hi, I’m trying to convey some information to the FBI about an identity theft fraud being perpetrated against me, and I’m a little puzzled by the information I got on a card given to me by a guard at the door of the FBI building on Sixth St., N.W. The card has an email address to be used to convey information, but the email address misspells the word “field,” and even if I use the email address with the spelling corrected, it still bounces back to me. So can you give me an email address please that works?

FBI: The FBI does not accept information by email.

AMG:  Well then why is there an email address on this card?

FBI: What is it about the card that makes you think that there is an identity theft problem?

AMG:  What?! The identity theft fraud is what I’m trying to report.

FBI:  Are you the victim of the identity theft?

AMG:  Yes, I am, which is why I am trying to report it.

FBI: Did someone put the card under your door or something?

AMG: No, as I just told you, I got the card from an FBI employee at the door of the FBI building on Sixth St., N.W. That was about 45 minutes ago.

FBI: The way to report information is to go online to http://tips.fbi.gov.

AMG:  I see that on the card, but I also see the phone number I used to call you. The card suggests that I can give you the information, as well. That’s why the number is 1-855-TELL-FBI.  Well, I want to tell the FBI.

FBI:  This number is only to report suspected terrorism or criminal activity.

AMG:  Isn’t an identity theft fraud that now spans four states an example of criminal activity?

FBI: The FBI needs evidence to consider criminal activity.

AMG:  Yes, well of course that is entirely reasonable, which is why I am trying to provide evidence to you.

FBI:  Well just go to the online address and provide it there.  [CLICK]

I did that. So now I have contacted my local police department, the FTC, the FBI, the IRS, the police department in Macomb, Illinois, the Division of Motor Vehicles of the State of Maryland to report the forged driver’s license, and finally, Geico, to report the insurance card acquired under false pretenses.

And the results? Geico cancelled the insurance policy at my request. The Maryland DMV called to assure me that my information is safe, and that the photo I sent them of the forged driver’s license was indeed that of a forged driver’s license. I heard back from the Macomb, Illinois police, informing me that the address I had given them was that of an assisted living facility, and was clearly not the domicile of any bad guys but just a typical part of the organized crime methodology that identity theft usually is. My local police department informed me that my case would not be turned over to a detective, since all they have is suspicion of attempted fraud in Montgomery County, and there are just way too many identity theft fraud reports to follow up. I have heard nothing whatsoever from any Federal agency: the IRS, the FTC, or, of course, the FBI.

In sum, someone out there is actively targeting me, trying to take advantage of and steal from me, and no law-enforcement agency, local, state, or Federal, is willing to do anything at all on my behalf. They all tell me it’s someone else’s responsibility, and so no one claims agency and it ends up being no one’s responsibility. I’m on my own, and so is virtually everyone else in my situation. As far as the FBI is concerned, it’s just as well. If they got involved, they might end up taking the side of the crook and accuse me of attempted identity theft.

Every so often I read some commentary in which the writer claims that Americans, relatively speaking, are under-taxed compared to the citizens of other advanced countries. Even in the past, when I have added up all the Federal deductions from my gross pay and from the small family business I help to manage, and then added state and local taxes to both personal and business accounts, I have not felt particularly under-taxed. But maybe that’s just me. Imagine how I feel now, given the “service” I have received lately from law enforcement? Hard to imagine that’s just me, given the rather low approval ratings for government these days and the tendency of most Americans to applaud any tax legislation that seems to lower their tax burden.

Now try to imagine the bereaved in Parkland, Florida, and how they feel about the services they have received lately from the Florida Department of Children and Families, their local police, and from the FBI. To add to their suffering, they are victims of taxation without competence. My problems are utterly negligible compared to theirs, I know. But the two problem sets are not entirely unrelated: It seems to always be someone else’s responsibility to assume genuine agency, whether within organizations or among them.

So, finally, let me ask this: Whose responsibility is it to connect the dots between the very large problems we face on account of dysfunctional government that cannot get bureaucracy to work properly and the often similarly generated small ones? If you know the answer, please tell me.