The procrastination solution you won't get around to reading

Rats, pigeons, and other vermin such as congressmen: all procrastinators. It’s likely that you, dear reader, are a procrastinator. Hell, you probably avoided reading this for a while. Paradoxically, procrastination causes considerable stress to the procrastinator; here is the Google result when searching for “dsm 5 procrastination”:

No, Google assures you, you’re not mentally disordered. But yet doesn’t it feel that way sometimes, when you just can’t manage to get things done? Here is the 2nd Google result:

OK, sure, the authors seem to say, it’s not a disorder BUT

Most people understand procrastination as unnecessarily delaying or failing to complete a behavior before a deadline. A lot of people procrastinate. Most people would call this a personality trait or tendency, while I’d call it a behavior, but let’s just agree to call it a habit. When habits or patterns of behavior are distressing, they can become pathologized, and clearly procrastination has that effect. People want to treat or solve procrastination – but personality traits are inherently resistant to treatment. After all, if you are a procrastinator, then procrastination is simultaneously a part of you and your own fault.

Then again, as all good behaviorists know, rats and pigeons procrastinate. OK, sure, Skinner didn’t call it that BUT just look at this:

In a mixed fixed interval/fixed ratio schedule, three organisms engage in low frequency responding until near the end of an interval, when their rate of responding increases sharply. One data path is a rat, one a pigeon, and one a monkey (Skinner doesn’t specify which is which). In a task where there is effort and a “deadline,” various organisms wait to expend the most effort until the very end. In this lens, are we talking about procrastination or efficiency?

Improbably this brings us to congress. In a truly classic article that you should absolutely read in full, Critchfield et al. use public information to compare the behavior of congress to lab animals. The information used is: 1) the length of the congressional session, which is a fixed interval, and 2) date of passage of bills, per session. Presumably, congresspeople want to pass bills before the end of the legislative session. Lo:

Critchfield et al. do some truly brilliant analysis (read it!!), but ultimately the evidence is clear: in an environment that involves effort and a deadline, scalloping-type patterns occur. 

At this point you might be wondering if the whole solution to procrastination is acceptance – just accept that this is a natural quality of all organisms. No, of course not! But also, yes, it’s step 1.

Step 1 of solving procrastination: accept that procrastination is a natural response to the environment. Acceptance doesn’t mean that you like it; it’s simply acknowledging reality. In cases where there are effortful responses with a deadline, responding follows a scalloped pattern. You, procrastinator, are not actually a procrastinator. You’re just like everyone else; in fact, you’re even like non-human organisms, such as congresspeople.

The good news about procrastination, from this perspective, is that it can be looked at as a schedule-induced (inductive) behavior rather than a personality trait. Alterations can be made such that procrastination is solved. It is not a part of you, in fact it is quite literally outside of you, as an environmental relation. 

The issue is easily conceptualized with an example: you need to write a report that is due in two weeks. Many people, if not most people, do little or no work on the assignment initially. The worst procrastinator might start on day 14 and work frantically into the night. A possible step 2 is setting more deadlines (note the work of semi-discredited researcher Dan Ariely is cited here). This is counterintuitive; if 1 deadline isn’t working, why add 13 more? Rather than completing 14 pages in a single desperate night, use rules (Ariely again!!) to evenly spread the work across 14 days of 1 page each. Engaging in behavior to affect future behavior is called precurrent behavior – examples could be carefully managing your calendar, putting your keys near the door (so you don’t forget them again, Stephanie!), or adding reminders to your smartphone. This involves engaging in more effort to set up the task, but as fluency increases the effort will decrease. It’s also one method Skinner used as he aged. “Distributed practice” should also improve the quality of the work output – studying, writing, second language acquisition, etc. are improved when distributed. 

Rules, used this way, will not work for all people all the time – in the linked study, about a third of the students procrastinated no matter what deadlines were imposed. Another possible step 2, in line with acceptance in step 1, is engaging in committed action according to your values (here are some steps). ACT interventions have been proven effective in so many ways that I refuse to cite anything to back up that assertion.

Choose your values, decide on behaviors that move you towards and away from your values, and commit to acting as such. But I’m going to do that later, you’re thinking.

Which brings us to step 3: data collection. It’s factually true that you, dear reader, have surely tracked data somewhere somehow someway. In fact, you might even be pretty good at it. However…for some reason this step – tracking your own behavior – inspires trepidation in everyone I talk to in real life. There is evidence that self-graphing alone can change behavior. Even Ernest Hemingway did it. That example is instructive, because you could slap a number on a piece of scrap cardboard – the format is less important than the behavior.

At this point you might be thinking: nothing will stop me from procrastinating and I’m not doing any of this. I don’t blame you! And I look forward to seeing you in congress. 

But if you think this entire premise is stupid, then you’re in good company: none other than Jack Michael agrees. A legendary figure in behavior analysis, in his 1980 ABA (later ABAI) presidential address he specifically describes attributing procrastination to scalloping as “superficial nonsense.” He continues: 

“This general approach is understandable in the case of the undergraduate student taking an introductory course in behavior analysis, and is reminiscent of such students' tendency to interpret industrial pay by the hour as a form of fixed interval reinforcement. But when such shallowness characterizes the verbal behavior of the professional behavior analyst, it is clearly bad news.” (p 15) 

For those of you who are not accustomed to the jargon of behavior analysts, he’s saying: u buffoon. I have to admit, coming across his comments at random and seeing that they describe ME was kind of like sitting inside a freshly purchased Tesla and watching the logo on the steering wheel slowly peel off to reveal “MORON TRAP tm” branding. Dang it!

At the same time, I think I see his point: the complications of human life are significant. To say that something like “study behavior” is due to a reinforcement contingency akin to a rat with a lever is simplifying things; in fact, even the rat with a lever is a simplification of true free operant contingencies for the rat. Even a rat has competing sources of reinforcement, etc. For humans, just on a superficial level, studying is more of a negative reinforcement contingency. The interplay of verbal rules, free operant contingencies, and mediating variables is less like a sharply defined stream of ABCs and more like a sloppy alphabet soup. It is simply not possible to take basic research and make a 1:1 comparison to human life.

Dick Malott makes a similar point at length in JOBM (with one of the most unhinged abstracts I’ve ever seen). At least, I think; this article could just as easily have been posted in the comments section of a blog post due to the pissy tone and barely comprehensible writing. (Behavior analysts used to have fun!) The point: researchers may take results and try to retrofit basic principles into a conceptual analysis (e.g., if employees become more productive in response to feedback, it was due to positive reinforcement). Explaining phenomena logically, but without supporting data, is sometimes known as a “just-so story.” In anthropology, this typically involves ascribing current behaviors as advantageous to our ancestors (e.g., if a study finds that “women be shopping,” it may be explained that cavewomen were the gatherers of berries, therefore…). One example in behavior analysis: behavioral momentum theory doesn’t explain all data, and in fact it is now understood to be a flawed metaphor at best.

And yet…the argument seems to boil down to, “life is more complicated than the lab.” It’s impossible to disagree with that assertion, but to put it bluntly: so? For a more erudite take, here’s Tom Critchfield, the co-author of the “scalloping in congress” article:

“[...] if our requirement for generalizing is that any environment strictly parallel any other, then I guess behavior principles have zero generality, because no two real-world environments are alike and most of them do not perfectly parallel lab schedules of reinforcement. As far as I know, we are the only science that tries to split this silly hair. Astronomers and physicists, for instance, try to account for complex motion of bodies in space based on principles derived from studying relatively simple motion. The fact that no lab study in physics has examined, say, the exact configuration of our solar system doesn't lead them to poo-poo the notion that gravity applies to the solar system.” (Personal correspondence, March 2024) 

With the right data, Critchfield says, we can make analogical claims – congressional workloads are scalloped like an FI schedule. With the right data, gravity has effects on galaxies (Aristotle believed that gravity was unique to earth while Plutarch surmised that it affected the universe).

At times, explaining complex interactions with basic terms bumps up against the bounds of reason. For example, describing a paycheck as positive reinforcement fails to understand both the nature of a paycheck and positive reinforcement. But science demands that we are able to conceptually explain successful intervention – how long would an effective medication stay on the market if nobody could explain the mechanism behind it? These mechanisms are sometimes poorly understood, or even entirely misunderstood at times, but that too is the province of science. 

In any case, I think Jack Michael gets it exactly wrong: I am not comparing humans to rats. I am comparing reinforcement schedules. I’ll allow Skinner the last words on the topic:

“I had the clue from Pavlov: control your conditions and you will see order.”