There is one correct way to amend a Constitution, part 2 of 2
What process amends a constitution while keeping society safe from autocracy? The answer is two majority votes, separated by some years. This is safer than requiring a super-majority in a single vote.
(This is part 2. See part one.)
How should we amend a Constitution?
Should we have a single vote that requires a super majority? In part one, we saw the problems that the nation of Hungary encountered, with a single super majority vote. In a crisis, when the public is in a panic, this approach can make too easy for an angry, populist party to gain power and then amend the Constitution. This is what happened in Hungary.
Should we amend the Constitution with multiple votes, and a super-majority? In part one, we saw the problems this caused in the USA. When it is too difficult to amend the Constitution, people will begin to sabotage it, as it no longer serves their real needs. They will find ways to cheat, ways to get around it, and then cheating will become the norm. This is what happened in the USA.
So what is the correct way to amend the Constitution? The answer must be two votes with a simple majority, in other words, two votes separated by a few years. But how many years should that be? Maybe 5? Or 10?
The myth of Katechon
Let's step away from reality for a moment, because sometimes an extreme fictional situation can offer a clarity that the real world lacks. We will talk about the nation of Katechon. This small island nation was supposed to be governed as a direct democracy, but a terrible plague had swept the nation leaving everyone with brain damage, so that no one could remember things for more than a few weeks. As such, political gatherings tended to bring together a mass of people who were overly perturbed by recent events, but unable to remember long-term concerns unless they were somehow reminded of those concerns by recent events. Gathered together these people amounted to an ignorant, angry, hysterical mob. Only one person was spared this terrible plague. Aware that a fully functioning brain was both rare and also useful in government, the people made this person the Chairman of the assembled crowd.
Can a government produce good laws if it is run by an ignorant, angry, hysterical mob that suffers from amnesia? Surprisingly, the answer can be a yes. Katechon had the happy rule that all laws had to be voted on twice, with the two votes separated by ten years. Thankfully the Chairman recalled this rule, and reminded the assembled public of it often.
An example of how this rule lead to good government:
One day a Brazilian person robbed some other citizen. The assembled public agreed this was an outrage. They immediately passed a law declaring that all Brazilians should be put to death.
A week later an Algerian person robbed some other citizen. The assembled public agreed this was an outrage. They immediately passed a law declaring that all Algerians should be put to death.
A week later a Norwegian person robbed some other citizen. The assembled public agreed this was an outrage. They immediately passed a law declaring that all Norwegians should be put to death.
A week later a citizen was driving to the assembly, when their car hit a huge pot hole. This was clearly an outrage. They proposed a law authorizing more spending on infrastructure. Other voters had also hit pot holes, so they all voted for the law.
A week later a citizen took their daughter to school where they noticed that school building was falling apart, with a broken heater and poor insulation and leaks in the roof. This was clearly an outrage. They proposed a law authorizing more spending on the schools. Others had also taken their children to school, and so, knowing the poor condition of the schools, they all voted for the law.
In every case the Chairman reminded them the laws could not take effect for ten years, and only if a second vote approved the law, at that future date.
Dutifully, when ten years had passed, the Chairman reminded them of the bills from ten years before. No one could remember why they'd been angry with Brazilians, Algerians, or Norwegians, so those bills were defeated at the second vote.
However, the situation with the roads had gotten even worse. The pot holes were everywhere and as people had driven to the assembly hall they had nearly destroyed their cars in the now gargantuan craters in the roads. The bill authorizing more infrastructure spending on the roads was approved overwhelmingly. And likewise, the situation with the schools had grown even worse, and everyone was now aware that their children were going to school in buildings that were falling apart. So the bill authorizing more spending for the schools was overwhelmingly approved.
In this way, an ignorant, angry, hysterical mob that suffers from amnesia can still pass good laws, because all of the bills that arise from short-term panics are eventually defeated, but those bills that deal with the public's long-term concerns, such as funding for the roads or the schools, successfully become laws, exactly because the concerns around these issues persist over time, and reassert themselves in everyday life.
I've mentioned before that I spent 20 years writing software, and my career informs my political thinking. In the world of software, developers must work with unreliable hardware that often burns out, plus unreliable software that is full of bugs, plus unreliable networks that often drop messages, and out of this mass of unreliable parts, a reliable system needs to be built. How is this done? With a large dose of redundancy. For instance, when sending messages over a network, software can send the same message several times, to be sure that it is received at least once. There are communication systems where 99% of all messages are lost, yet the messages are still reliably delivered, because at least one message is delivered, after multiple attempts. A similar strategy can work to ensure that good laws come out of terrible governments. Multiple votes is a kind of redundancy, a filter that limits the spam that might otherwise become law.
The people of Katechon wait ten years to finalize a law. With this simple rule in place, the worst government in the world ends up passing laws that are good for all of society. Even though 99% of all the laws proposed are both stupid and harmful, the laws that get passed tend to be beneficial, because they arise from the long-term needs of the people.
Every society needs to find a balance between stability and flexibility
Again, as we said yesterday in part one of this essay, every society needs to find a balance between stability and flexibility; every society needs to adapt to change while keeping the important fundamentals solid. We should actually crave for the law to change as fast as possible, provided three conditions are met:
1.) that we know the issue is non-trivial (important to a large number of people)
2.) that we know the law does not arise from panic or vengeance or any other hot emotion
3.) that we know that the bill itself was vetted by many people over a long period, so that it is technically strong, containing all necessary qualifiers
Would it make sense for Katechon to also have a rule saying that laws should be passed by a super-majority? Clearly, no, such a rule would be harmful. If an issue can maintain majority support for ten years, then the issue has already proven that it is a long-term concern of some importance. It is not trivial, and it does not arise from some temporary panic. It is an issue that must be addressed. Demanding a super-majority, after so many years, would only make the system inflexible and brittle. Requiring a super-majority, after so much time had passed, would again make it too difficult to change the law, and therefore we would again be facing the problems that the USA has faced, which we analyzed in part one.
But is ten years the ideal amount of time to wait, before passing a law?
Every successful society finds some way to put a bill through multiple affirmations before allowing it to become a law. According to Herodotus, when the ancient Persians were on the rise to greatness, they had a rule that all proposals had to be considered twice, once when sober and once while drunk. Presumably this habit arose after too many laws were passed while the gathered nobles were in a drunken stupor, and these laws turned out to be less than ideal. Whatever the origins of this habit of voting twice, the practice of having more than one vote is the first and most important step that any assembly can take if it's going to pass good laws. Likewise, most Western democracies currently insist on multiple steps before a bill becomes law. Even in countries where the legislature is unicameral, there are often procedural rules in place that force multiple readings of the law before it can be voted on.
And yet, there will sometimes be a real crisis, such as war, which will demand a fast response. So “two votes, separated by 10 years” is not really ideal. We need something more flexible than that.
The size of the majority on a first vote should indicate how soon the second vote is to be held
Let's imagine what it would mean if it took ten years to push through an amendment. That would have offered some protection to Hungary -- Fidesz gained power in 2010 and pushed through all of their amendments immediately, and presented this vast change to the EU as a fait acompli, but if they had to wait till 2020, they would have been attempting to push the changes through after the EU had time to state its objections, and after the Hungarian public had ten years to consider what such changes really meant. A ten year wait does not offer perfect protection against tyranny, but it is still a potent tool for blocking demagogues who want to use a specific panic to consolidate their hold on power. Fidesz now hangs onto power because it controls all of the media, but it gained control of the media because it changed the constitution in ways that let it drive out all foreign news outlets, and local news outlets were blackmailed into selling out to oligarchs loyal to Fidesz. If Fidesz had been unable to amend the constitution, it never would have gained control of the media, and so it would have faced real scrutiny, and so it would probably less popular right now.
Suppose that in the USA a Constitutional amendment could be passed by two majority votes in the Senate. FDR's National Industrial Recovery Act of 1933 could have been allowed by the Constitution by 1943. In the case of the Great Depression, the delay of ten years would have been painful, but perhaps less painful than the current situation, where more than 85 years have passed, and the expanded powers of the Federal government still have no justification in the Constitution, at least none that can be discovered by a good-faith reading of the written Constitution. Bad faith readings of the Constitution now corrupt the purpose of the high court.
All the same, we must admit, specifying some fixed number of years is not perfect. What we actually should want is two votes, separated by a certain number of years, with the number of years shortening based on the strength of the support for the amendment. In a place like the USA Senate, which has 100 members, we could have a rule that the second vote must occur zero years in the future, plus one year for every 4 senators who voted against the amendment during the first vote. So assume the amendment is highly controversial, and during the first vote it passes with a vote of 51 to 49. The math is:
49 / 4 = 12.25
Let's specify a rule that we round that down to 12. So the Senate needs to wait for 12 years before taking the second vote, to see if that amendment becomes part of the Constitution. For such a controversial amendment, 12 years offers plenty of time for the opinion of the public to shift, either in favor of the bill, or against it.
But what if a major war breaks out, and suddenly the government needs a new power? Perhaps the government finds that it needs strict control over the Internet, or needs to limit discussion of some new technology. Such an expansion of government would need a Constitutional amendment. In the face of a true national crisis, it might be possible to get at least 97 senators to vote for the amendment -- in that case, the second vote can be taken instantly, and the amendment therefore be passed instantly.
All of which is to say, the ideal process for amending a Constitution has to be faster than the current USA system, but slower than the Hungarian system, and in normal circumstances it should take a few years, but there should be a mechanism that allows for instantaneous changes if everyone in the legislature agrees that the nation is facing a true crisis.
How would our flexible rule apply to our two examples of Hungary and the USA?
After the elections of 2010, Fidesz held 70% of Parliament. If we translate our previous rule to this context, we could say that instead of every 4 senators, the rule applies to 4% of the vote. Assuming most of the votes for Constitutional change would have been 70% to 30%, then:
30 / 4 = 7.5
so Fidesz would have to wait 7 years before pushing through their authoritarian takeover. As said above, this would give the Hungarian public more time to re-think the issue, as well as more time to recover from the shock of the 2008 financial crisis. After 7 years, the public would see what kind of normal laws Fidesz had passed, and that has, in real life, tempered the public's trust of Fidesz. It is unlikely that Fidesz could have carried out their constitutional coup if the EU and the Hungarian public had been given 7 years to respond.
In the USA, the National Industrial Recovery Act of 1933 passed the house with 81% of the vote. Following the rule above, with 19% in opposing the act:
19 / 4 = 4.75
So the USA government would have needed to wait 4 years before being able to confirm that the Constitution was amended in a manner sufficient to support such an act. That is, the law would finally be Constitutionally legal in 1937. Considering that FDR did not finally have a compliant Supreme Court until 1937, in our scenario the NIRA would have been legal at the same time as it was legal in real life, however it would have been legal because the Constitution was properly amended to make it legal. No bad faith interpretation of the Constitution would have been necessary, no blackmail of the Supreme Court, no threats of court packing. A beautiful and democratic process would have happened:
1. the government needed a new power
2. the government followed the official, formal process for amending the Constitution and so granted itself this new power
3. the government then rescued the nation from the Great Depression, thanks to the new power granted by the amended Constitution
To summarize the whole argument:
1. Why should we want Constitutional change to happen reasonably fast?
Because if the Constitution is too difficult to change, then people will begin to find ways to sabotage it, and also the sense of unresponsive stagnation only serves to bring the political system into contempt. (In part one, the example given was about marijuana.)
2. What is the appropriate limit on the speed with which the Constitution is amended?
We must ensure the decision is not driven by rage, panic, or any other hot emotion, but when we can filter out such bad decisions, we should want to see every amendment, if supported by the majority, to pass as quickly as possible.
3. When should the courts amend the Constitution?
Again, we quote Judge Richard Posner:
“Perhaps the courts are authorized to plug at least the most glaring gaps... We find it reassuring to think that the courts stand between us and legislative tyranny even if a particular form of tyranny was not foreseen and expressly forbidden by framers of the Constitution.”
The courts are meant to be independent and non-political, and so they are corrupted when they are granted legislative powers. If they must plug "the most glaring gaps" then the system works best when those gaps are as small as possible, and that can only happen when the legislature is able to pass amendments with reasonable speed.
Real freedom can only happen in a society that has an accountable political system, with a clear process for change, guided by the rule of law, and protected by independent courts. We've seen that such a system was destroyed in Hungary, and we can plainly see how such a system might be destroyed in the USA. Such a system is best protected when the Constitution can only be amended at a moderate pace, slower than what is allowed in Hungary and faster than what is allowed in the USA.
At this point, you might reasonable ask, "What is the point of this essay? Is there any chance any such system of Constitutional amendment will ever be enacted in real life?"
I don't know the answer to that, but I'll say this: democracy is currently in crisis. Authoritarian demagogues have gained power all over the globe. Putin in Russia, Orban in Hungary, Bolsonaro in Brazil, Modi in India, Duterte in the Philippines, Erdoğan in Turkey, Kaczynski in Poland, and, the most dangerous of all, Trump in the USA. Only by a major mobilization of the public was Trump pushed out after one term and if he returns it is uncertain whether democracy will survive in the USA.
How does this era end? Will there be fundamental changes to the ways that democracies operate? None of us know the future, but I suspect many changes will be necessary before democracy, as a form of government, can again feel like it has the momentum of the world behind it. I offer this essay to suggest one of the changes that will need to happen: careful thought should be given to the speed at which Constitutions are amended.