50 years ago, a researcher named Walter Mischel and his colleagues ran a series of experiments that are now often referred to as the Marshmallow Test. The basic premise was to measure children’s “self-control” by examining whether and how long they were willing to wait to receive a desirable consequence, which was commonly a cookie or candy or – less frequently, a marshmallow. A child would be brought into a room and offered one marshmallow that they could eat as soon as they wanted. BUT they are told that if they instead choose to wait until the experimenter returns, they will receive two marshmallows. The critical variable of interest is determining just how long a person will wait for that second marshmallow.
Several fruitful findings have come from this initial work. For one, there is tremendous individual variability in ability to wait – even among relatively similar populations. With the 4-5-year-old children, about a third of them were able to wait the required 15 minutes for the second marshmallow, but most who didn’t wait consumed the first within 30 seconds. Another important finding is that there are several correlations between how long children wait and later-life measures such as higher SAT scores, lower body mass index, lower money mismanagement problems, and fewer substance abuse problems. Given the societal importance placed on these types of outcomes, you’d be absolutely correct if you assumed that this has prompted many, MANY additional studies. As a research area, it’s essentially gone viral.
Though I don’t want to imply that a novel phenomenon has recently been uncovered here. The human debate between wanting something bigger but then sometimes choosing something less meaningful… all because we can have the less meaningful thing faster, has been at the very core of many political or religious debates and can be found sprinkled in literary works for centuries. It’s a classic human problem. It affects not only our marshmallow choices but our ability to plan for retirement or stay in committed relationships, among other everyday mundane behaviors and more significant larger life decisions.
Using drugs is one pattern of behavior that is at once seemingly every day mundane yet longer-term significant. It has been a key clinical area of study for researchers interested in this classic dilemma – the dilemma between our impulses and self-control. Impulsivity has long been a defining feature of several clinical disorders – including substance use disorder, and behavioral researchers realized that impulsivity can be studied in a more precise quantitative manner using tools derived from behavioral economics. In a manner very similar to the marshmallow task, people can be asked to choose between commodities differing in size and delay to establish how long they’re willing to wait for the bigger consequence. For example, if you were to be tested, you might be asked questions such as: would you prefer to receive $25 right now, or $100 in a day? What about $25 right now, or $100 in a month? $25 now or $100 in a year? If you’re like many people, you’d initially opt to have the larger amount in a day or even a few days or weeks. But when the delay to the larger amount increases, the “best” choice becomes less clear to us. Amount and delay are powerful variables that impact our choice behavior, and when their influences converge it can become quite challenging for us to make up our minds so to speak.
When substance abusers are asked these types of questions, their choices are remarkably different as compared to matched controls (i.e., people who do not use drugs but are otherwise as similar as possible). In the 1990’s, Greg Madden, Warren Bickel, and their colleagues first examined this with opioid users, finding that addicts were much more impulsive (less willing to wait for larger, more delayed consequences) than their counterparts. Not just a little more impulsive, either – but significantly more impulsive. And this finding is not limited to heroin users. Since this initial study out of Bickel’s lab, delay discounting has been examined in users of just about every drug (e.g., cigarettes, alcohol, crack/cocaine) and also in other clinically impulsive populations (e.g., pathological gamblers, children with attention-deficit hyperactivity disorder). With few exceptions, substance users are less willing to wait for delayed rewards than are the non-abusing population.
So – it’s a robust finding. But what does it mean, practically speaking?
One is that when a person places a very high value on consequences that are available soon, and devalues those that are more remote, this can make it extremely challenging to stop using drugs. Furthermore, it can theoretically make relapse a more likely scenario for someone who has managed to quit.
Consider this: the value of using drugs is already quite high to someone who is addicted. As I discussed in an earlier blog, while there are some positively reinforcing effects of substance use (such as a particular drug effect/feeling you might get), a lot of the consequences of drug use are negatively reinforcing – eliminating the unpleasant withdrawal effects. This negative reinforcer can be VERY powerful. Honestly if it weren’t very powerful, we would never encounter people who want to quit but are having a difficult time (which, of course, we do). So, we’re starting with a consequence – using the drug – that already carries a good deal of weight on its own. If it’s available immediately or at least relatively quickly – all the more powerful it becomes. For someone who is then considering using a drug versus abstaining, it’s a sad reality that many of the positive consequences tied to abstaining are delayed. Not only are they delayed, they are at best probabilistic. For example, abstaining might theoretically lead to a better life in many regards – increased income to spend on other things, better relationships with those around you, more solid performance at your place of employment, better long-term health prospects, etc. All of these consequences are likely, but unfortunately, they are delayed and not even a guarantee for someone who is considering abstaining.
Now, let’s add more insult to injury. On TOP of the aforementioned issue of pitting a highly valued relatively immediately available reinforcer of drug use against a highly valued theoretically possible but delayed and probabilistic “better life” outcome, the delay discounting research that I mentioned has repeatedly shown that substance users strongly devalue delayed outcomes. Meaning that when given the choice between a smaller yet sooner outcome versus a larger yet delayed outcome, they’re more likely than their peers to take the small one. More likely to take the $25 now versus $100 in a month. More likely to use drugs now versus abstain and wait to reap the larger life rewards.
There’s a lot more that researchers have added to this story, but frankly – a good deal of the very practical implications of this remain unclear to this day. This is one of the reasons this line of research has blown up over the past decade. Here are some of the important questions we’re still learning about:
This research has largely been correlational in nature – what are the causal variable(s) here? Does substance abuse cause impulsivity, or is it the other way around? What are some of the neural correlates of delay discounting? Is our ability to wait longer for delayed consequences hereditary, and could this account for correlations of substance abuse among family members? And probably the most important question of them all: can we do anything to change our underlying impulsivity? If so… what?
I’m going to focus in on these issues in some upcoming blogs – subscribe to stay tuned!