Simplistic Solutions Can Have Unintended Consequences
An editorial written by an American Airlines pilot in the Fort Worth Star-Telegram caught my attention (“Unintended consequences of the Passenger Bill of Rights“). The pilot makes the case that the actions of Congress taken to prevent multi-hour passenger delays on the tarmac give incentives to airlines to proactively cancel flights rather than face the risk of million-dollar fines for a single flight.
This “Bill of Rights” is arguably a simplistic overreaction to a problem, one that didn't get to the root cause of the original problem.
The pilot writes:
The Passenger Bill of Rights is the wrong answer to the right question that demonstrates two important points. First, a simplistic legislative solution is completely inadequate to a complex problem like tarmac delays.
And second, for all who lobbied for this legislation based on a handful of overpublicized and anecdotally enlarged tarmac tales, when you're in line waiting to rebook your travel, remember that you got what you asked for: You're not waiting on the tarmac. You're simply not going anywhere.
Well-intended actions often lead to unintended consequences. This is a core lesson of the “System Dynamics” field popularized by Peter Senge and his bookThe Fifth Discipline: The Art & Practice of The Learning Organization. I was fortunate to take a course in this during my graduate studies at MIT. We learn many lessons of simple actions that, while locally helpful, make the system worse.
One classic example was towns along the Mississippi River building levees to keep flood waters back. As our professor said, your town's levee only had to be an inch taller than the levee of the town across the river. This led to a levee arms race that inevitably pushed flooding further upstream. The further upstream, the worse the flooding. Locally brilliant, globally suboptimal.
In the case of the airline, they framed the problem as “passengers shouldn't have to wait without food, water, or working toilets.” Surely, there was a solution to be found to provide water, food, and toilets to passengers – bringing it to them or having better processes for bringing a plane back to the gate temporarily. Since the airlines couldn't create a compelling and workable countermeasure, Congress stepped in with a “solution” that sounded great as a sound bite. But now you might be more likely to miss the start of a vacation altogether instead of being delayed for three hours. Thanks, federal government!
The behavior sometimes occurs in misguided simplistic “solutions” in manufacturing. During my grad school internship, the company was struggling to recover from a misguided “just in time” effort. If inventory is a cost and Lean says (or so they thought) that you should strive for zero inventories, they cleaned out the “finished goods” inventory that was really an input for final product of an expensive product for which they could sell anything they made.
But, the upstream silo thought it was making its own silo's budget look good. They had a production process with long lead times and highly variable quality yield rates. These were far from the pre-requisite conditions for JIT. So, what should have been predictable was that while inventory was low, the unintended consequence was that they couldn't make planned delivery to the internal customer. The final production area screamed about how they didn't have a critical component needed to make a $15,000 sale (lost business was often taken by an equally capable competing product).
We can see similar overly-simple thinking in healthcare. In the UK, a culture that loves arbitrary targets (something Dr. Deming would have disapproved of, I'm sure), the problem was seen as patients waiting too long in the “Accident & Emergency” department. So the government set a target of an arbitrary 4-hour limit for waiting in A&E.
Hospitals responded, in many cases, with dysfunctional behavior – unintended consequences. Ambulances were kept parked just outside the door with the patient still inside so that, by a technicality, the 4-hour clock was not ticking yet. Or, patients get admitted unnecessarily (they still haven't seen a doctor, perhaps), tying up a bed that is needed more by another patient who is now waiting (but hasn't hit their 4-hour limit yet). Instead of improving the overall flow, a simple “solution” was created that didn't take the whole system into account.
Whether it's an airline, a factory, or a hospital, we need to combine Lean with a Systems Dynamics view of our work and value streams. Instead of local solutions that harm the whole, we need to avoid the “quick fix” and the “easy answer” that might cause more harm than good.
As we move forward with the Passengers' Bill of Rights, here's a chance to see if the government practices PDCA (Plan-Do-Check-Act) – if the law isn't working (or the unintended consequences are worse than the benefit), Congress should kill the law. Now I'm being overly simplistic, eh?
Note: After originally drafting this blog post for future publication, the WSJ had an article on June 3 (“The New Rules Of Airport Delays“) that said the airlines DO have some plans other than just canceling flights, including:
Airlines and airports have elaborate plans, stocking up on bottled water and pretzels at some airports, even deploying new buses, vans and “drive-by gates” to quickly unload passengers who want to get off long-delayed flights. Foremost among the changes, airlines say they will cancel more flights to lessen the chances that planes will sit without taking off or finding an open gate.
But it seems preemptive cancellations might still be a problem.