Laboratories of Democracy: Federal Standards

Federal (nation-wide) regulations offer many benefits: especially uniformity & stability.
  • Uniformity – I mean, really, what large-scale company/nonprofit wants to keep track of 50 different sets of state-level regulations!
  • Stability – If you are fearful that regulations will change, that uncertainty causes hesitation and slows economic activity.  One set of federal regulations would be far more stable (unchanging) than 50 state-level sets of regulation.
But of course, if there is one set of federal regulations then any flaws in it effect everyone – and are harder to revoke than state or local regulations.
Possible Solution:
Short version – Allow each state to experiment, but over time require states to adopt the most “successful” of the experiments so that eventually the entire country is on a uniform regime.

Details – When the national legislature determines that a given area of law/regulation should be unified across the nation, congress follows this process:
  1. Design a uniform “national metric” by which each state governments’ regulatory regime will be evaluated.  This national metric is made public.  For X years, each state is its own laboratory of democracy, trying out whatever system it wishes (or continuing/tweaking the system it already had).
  2. After X years, each state’s regulatory regime is evaluated according to the national metric.  The regimes in the top Y% are put onto a Choice List.  Each state legislature then must adopt one of the regulatory regimes from the Choice List.
  3. In another X years, all states must adopt the highest scoring of the remaining regimes.
If this unified regime requires adjusting in the future, the same process (or an abbreviated one) could be used.
  • Congress decides we need to cut greenhouse gas emissions, but without killing economic growth.  They create a national metric that represents a tenuous consensus on how best to measure the desired result.
  • At the start of year 6, Congress publishes its Choice List, detailing the top 5 states’ regimes and the National Metric scores.  ME=80, CA=72, WA=64, MA=60, RI=58.
  • Each state legislature votes to adopt one of the 10 Choice List regimes. 20 states go with the ME plan, 15 with CA, 10 with WA, 2 with MA, and 3 with RI.
  • At the start of year 11, the scores come in as follows: ME=82, CA=80, WA=89, MA=71, RI=50.  As you can see, most regimes improved their performance over the prior years, however, they did not do so equally!  WA went from the 3rd ranked option to the 1st ranked.  As such, in year 12, every state is required to adopt WA’s regulatory regime.
How This Addresses the Need:
  • In 2X years the nation has a clear, unified regulatory regime that everyone must abide by. No complications from 50 different sets of laws, relative stability of the regime is in place.
  • The process by which this regime was chosen is (I think) less susceptible to the large scale political-only influences and allows the 50 “laboratories of democracy” to have their try, but only the best system was allowed to survive.  Also, the “one big mistake hits everyone” will occur less frequently/severely because each system is put to the test.

What do you think?


10 thoughts on “Laboratories of Democracy: Federal Standards

  1. What if solution “A” works great in 30 of the 50 states, but would be disastrous if applied to the other 20? Are they still forced to convert?What is applying solution “A” to state-#1 helps it but hurts state-#2? E.g. the shutting down of all coal usage bankrupts West Virginia, Ohio, Kentucky, and Illinois.

  2. Tom, excellent point. Of course, our existing system has this flaw too! That said, I agree that my proposal would be worse in this regard. At first I was thinking you could create rules for how these regulatory regimes have to be set up. Example: You’re not allowed to say “All factories outside of our state must pay twice the greenhouse tax as factories in-state” – but that kind of rule system lends itself to avoidance easily. I can’t think of a way around the problem – can anyone else?Alternatively, you could change my original proposal around a bit to be something like this…Step 1 stays the same, except the national metric also has a scaling factor (probably non-linear) for the population living under the regime.Step 2 changes – now the national legislature must select a regime for the entire nation, but it can only select from a Choice List of the top Y% of regimes.This change ensures that the usual checks-and-balances of state legislator votes minimize the chance of a policy that will gravely injure some states – yet they are forced to pick from a list of regimes that have survived a contest.

  3. Or…Step 1: same w/ metric adjusting for populationStep 2: same as original proposal, but states may change their regime as many times as they want so long as they pick from within the Choice List.Step 3: same as original proposalBy incorporating the population scaling factor, and by allowing states to change their regime as often as they like, I think this might offer “protection” at least as good as what we have now.

  4. To assume the federal gov. is less corruptible then the local gov. I think is dangerous. If anything I think it’s the other way around because they have more power and influence. What if each town could adopt the laws of any state? Then folks who love the laws by a state could then elect them as their own. If we could remove geography as a constraint, I think it would make it would empower local citizens because a whole town could elect to pay taxes to a different state if they did not like what their state was doing.

  5. I think you’ve just moved the problem from “depend on centralized bureaucrats to create good regulations” to “depend on centralized bureaucrats to create good national metrics.”MAYBE that would be better… but I doubt it. There would be immense pressure to create metrics that benefited certain industries or interest groups.Also, what incentive would the states have to strive for a great National Metric score? If there’s a natural incentive (if the state’s citizens want it), then it should happen naturally– no need for the feds to get involved.If not… then they’ll simply do the cheapest possible thing, regardless of effectiveness.

  6. The adjustment period of the proposal is unclear. The trouble is, by the time you figure out what works well, it’s time to create a new system. Things change too quickly for this to be effective. So say you find a way to cut emissions effectively in one state in Year 1. By Year 5 it’s clear that they’ve done the best job in cutting emissions. But at Year 4.5, nifty new technology comes out that’s twice as good. Rather than having the flexibility to adopt the new system, everyone’s forced to use an archaic system. And suddenly, you’ve lost a fertile testing ground.

  7. RE Gavin’s comments…I can easily convince myself that asking federal legislatures to create a metric is not significantly more or less political than what they do in the real world anyway. The benefit comes from the different systems then having to compete in a clearly measurable way.AS for incentives for the state legislatures – that’s an area I don’t have a good response for. The only thing I can say is “No state wants to switch to using someone else’s system, so if they have the best then they get to keep their current system.”What sort of incentive structure would you recommend (if any)?

  8. RE Liz's (AKA CSA's) comment:You make a great point, but the system we have now has the exact same flaw. I think you can mitigate this kind of issue if you remove technology specific regulation. Example: "You must put device X on your smoke stack" Such regulations go out of style once a better technique or technology comes around. However, if you design your regulations around numerical requirements and don't care HOW someone accomplished it (within existing legal/ethical/safety/etc law). Cap & Trade is better than "device X". A CO2 tax is better than a device-specific regulation, etc. Ideally you could design the metrics to be technology agnostic – but that might be getting too dreamy :).

  9. I like the ‘try to measure what you’re doing’ aspect a lot, by the way. The process for making code run faster is:+ Measure it+ Change exactly one thing+ Measure it again: if faster: keep the change, if not, undo the change.I’d like to see regulations work the same way; they should be automatically repealed if they don’t have the intended effect. Of course, we’re regulating such tiny risks these days (like second-hand cigarette smoke in restaurants) that actually measuring the regulation’s effects might be impossible.I don’t know how to align the incentives of the bureaucrats with what is “good” — I suppose rewarding the State with the best National Metric score with extra federal dollars would work.I agree with Liz’s point; change happens.

  10. To my mind, "laboratories of democracy" is a hit-or-miss proposition. It's great to experiment when you have something that lends itself to a "right" answer, and not to put too fine a point on it, but we may be in the midst of one of those processes with respect to gay marriage (what's the count today?).Tom's first post indicated a clear qualitative example of a lack of a right answer. If I read the post right, the issue is that the feds were going to come in and regulate anyway, so at some level a unified system is a given.So to combine Tom's and Gavin's ideas, the idea of a unified metric may be inherently biased by the fact that you're using a state-to-state experimental design. For instance, at the early stages, a particular design may be tied to a poor set of starting conditions (which because regulations are so inherently multivariate, would be difficult to correct for). Or, maybe there's a very real possibility that the experimental standout isn't necessarily the best steady-state solution. Even with an iterative process, some elements of competition – for instance commitments made by entities like businesses at earlier experimental levels (like where to build a plant) would always be present up until the moment a final plan is adopted.But Gavin's point I think is best taken with the idea of a single metric being troublesome. Sometimes issues like these have moral implications: How is "worker safety" going to be balanced against "cost savings?" – and for political reasons, I promise you it's not the actuarial approach taken by Ford.While I'm quick to decry politicians for not making knowledge-/metric-based management decisions, sadly it's because giving them even a few facts is sometimes too many. 😛

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s