Goodhart’s Law and Addiction
Jul07

Goodhart’s Law and Addiction

In a previous post I wrote about Goodhart’s Law in relation to prohibition. I want to give another example of Goodhart’s Law with addiction, this time with opioids. In a recent report making the national media rounds, Senators have written a letter to CMS, the government agency responsible for Medicaid and Medicare. The letter states that there is growing anecdotal evidence that surveys regarding pain scores are causing doctors to prescribe more opioids. Well, yeah, but this is Goodhart’s Law. “When a measure becomes a policy, it ceases to be a good measure.” CMS measured pain scores. They wanted to improve pain scores. So they instituted the policy that pain scores should go down. Pain scores ceased to be a good measure of quality of care, and instead became the driving force for increased prescription of opioids. By the way the same thing is happening with patient satisfaction scores and the intimidation of doctors who actually feel they’ll lose their jobs if opiate seeking patients write bad reviews of them. The real problem isn’t opioids or pain. The real problem here is the system’s response to the problem. Without ever trying to find the root cause of the problem, they just declared, “Solve the problem.” You just get another problem. I’ve written before about TOC and its usefulness in addiction here and here. This is another good example of where TOC could help. Instead of fixing what the problem looks like on the surface, we could use the TOC thinking processes to find the underlying common cause. We can then plan out how to affect the change and foresee the negative outcomes. Before implementing, we could address the negative outcomes and create an even better plan for change. But that, unfortunately, is not how large organizations like government work. And it is unfortunate, because where TOC is used we see remarkable change very quickly. I would like to see addiction go away tomorrow, and while even that isn’t possible with the TOC thinking processes, using them would make it happen a lot faster than it’s happening...

Read More

Goodhart’s Law and Prohibition

Goodhart’s Law: Any measure that becomes the object of policy ceases to be a good measure. Goodhart’s is a universal law. It essentially says that if you measure outcomes and focus on those outcomes as the point of the exercise, then they’ll cease to be good numbers. Instead, one should measure outcomes and then focus on changes to the system that do not include the outcome. If one focuses on changing the outcome, it will no longer be a useful outcome to measure. Examples of this can be found at all levels For instance, a couple of decades ago American Medicine and the Joint Commission began to focus on pain as an outcome. Practices were established to measure pain and then decrease pain. The point of most of the efforts was not to find and negate the source of the pain, but rather to decrease pain itself. Now, two and a half decades later, we have an opioid overdose epidemic and measured levels of pain in our society that are higher than 25 years ago. This, in spite of good evidence that opioids don’t work for pain for more than about 3 months. We are so focused on the outcome that we cannot see the system that brings it about. Let’s take another such outcome, drug use. Many have noticed that there are rising levels of drug use in our society. The War on Drugs has been a major response to that. It was not a War on the Causes of Drug Use in America; it was a War on Drugs, a war on the outcome. Our government has been measuring the outcome since that time, and it’s only getting worse. So what is the alternative suggested by Goodhart’s law, and why do so few people choose it? The alternative is to measure the outcome and then consciously not attempt to change it. Rather, one looks at the system that produces it and attempts to form hypotheses about how the system brings about the outcome. Then one would change the system and remeasure the outcome to gauge the effectiveness of the change. Unfortunately, this method usually includes things one would rather not look at. So, instead we focus on the outcome. Too many people wanting to smoke pot? We’ll make it illegal and then less people will want to smoke it. That didn’t work? Oh well, we’ll increase the penalties and then less people will want to smoke it. That didn’t work? Oh well, poison the crops so that anyone smoking it will go blind. That didn’t work? Oh well, we’ll increase border protection to keep it from coming...

Read More
Did Prohibition Work?
Jul30

Did Prohibition Work?

In the last week, and in the context the New York Times coming out in favor of repeal cannabis prohibition, two of my colleagues have used a graph that I showed them from my book to make the point that prohibition worked. Here’s the graph: So yes, in the year after prohibition, this country consumed half as much alcohol for every man, woman, and child as it did in the year before prohibition. Both of these colleagues used this data to argue for prohibition of a drug, in this case cannabis. They missed the larger point of the graphs that follow in my book. First look at the graph of alcohol use. It was pretty steady at around 25 gallons before prohibition, but then started to fall off even before the law was passed. While the end of prohibition shows that the per capita intake was only 10 gallons, half of the 20 gallons per capita before prohibition, notice how fast the number rises. The slope was much higher than seen before prohibition. We see the same increase in craving when someone with addiction stops using but doesn’t get treatment. Periods of “stopping” often end with use escalated to levels above those seen before the stop. Yes alcohol use dropped, so will use of any drug when we make it more costly, either socially or economically, to use. And normal people who don’t need or generally use drugs will probably not use it at all once we make it illegal. However people with addiction will still look for something to calm the symptoms of the biological brain illness they suffer from. So, what, I asked myself, might addicts who used alcohol have moved to when alcohol became illegal. Here’s the graph of cigarettes. In that last year before prohibition, 1919, America drank 19.6 gallons per 10 people, or 1.96 gallons per capita and smoked 727 cigarettes per person. Let’s just assume that we’re getting dopamine from both and multiply them together to get what you might term a “dopamine load.” 1.96 X 727 = 1424. Now let’s look after prohibition. In 1934, America drank only .97 gallons of alcohol per capita and smoked 1483 cigarettes per person. .97 X 1483 = 1438. No decrease; those who needed dopamine and couldn’t get alcohol just smoked more. Here’s a graph of the combined product for each year: One colleague answered the New York Times in a letter to the editor and said that while prohibition may have been a political failure, it was a public health success. Really? Before prohibition, alcohol use was falling. After prohibition it rose faster. Before prohibition...

Read More