More thoughts on the TC rule
I’ve been hugely busy for the last 2 weeks. When I wrote out the rule Friday morning for the first time, I didn’t have any time to write out reasoning or even about the choice of targets.
I am still very busy, but I have a little bit of additional background on the choices, plus some very quick notes.
I’ve realized this is more of a functional finance rule than an MMT rule. MMT has a rule already for this – it is the famous and controversial Employer of Last resort. Famous in the MMT world, but unfortunately controversial for most people. Part of the reason I created a rule was to form a practical rule we can use today, without the need to implement the ELR immediately. Let’s face it, the ELR won’t be voted into law next year. We need something useable and practical now. Thanks to Scott and Beowulf for pointing out the rule is more functional finance than MMT.
Since a country with a sovereign currency cannot go bankrupt, the only remaining worries for this government is the inflation rate and the unemployment rate. I’ve gone into this with some detail here and over at Interfluidity.
It is very possible that Interfluidity is correct and we need to be concerned with a discontinuous jump in inflation. I don’t think this is the case for reasons I think I laid out in the comments section at Interfluidity, but it is entirely possible (and thinking about Soros-Popper, extremely likely) that I am incorrect. We need more research into inflation under a sovereign currency regime.
So we have two targets – the Feds famous dual mandate – of unemployment and inflation. I prefer to target them directly rather than indirectly through GDP. However, Scott Sumner and David Beckworth have made a good case for targeting NGDP. This targets inflation and real growth in a single easy to understand number. I consider NGDP targeting to be a superior to than the Taylor rule.
Still, I don’t think an NGDP target is as good as targeting unemployment. Unemployment is bad for reasons beyond the sheer economic loss associated with able people sitting around. We do not need to make a choice between everybody working and high economic growth – we can presumably have both. NGDP could theoretically limit the economy below full employment due to assuming growth causes inflation. An unemployment target is less likely to run into a self imposed limitation on economic growth.
Warren Mosler has hinted that he thinks that real economic growth could be 10%+ with good policy. China has grown at this rate for decades, and Japan had growth approaching this level for decades until they changed their policy. We don’t know what is possible for long term growth with good policy – we have not yet had the good policy part. An NGDP target is a potential self-imposed throttle on economic growth.
Beowulf pointed me to Pavlina R. Tcherneva’s paper on fiscal policy. In it, she makes a great case for targeting effective demand over aggregate demand. The ELR is superior to the TC rule in that it targets effective demand, where the TC rule only targets aggregate demand through the mechanism of % government deficit. It is a must read paper on the topic, and I wish I had read it the day it came out. The distinction between Aggregate Demand and Effective Demand eviscerates NGDP targeting or targeting inflation alone. I cannot recommend this paper highly enough.
Beowulf points out that Okun’s law has been unstable. Yes, it has. As far as I know, most economic regressions are unstable. It’s a problem with any rule that uses empirical relationships in with economic data. Still, I think using the current accepted multiplier of 1.8 would be a good place to start. [Update: Beowulf points out why not just use Bernanke's 2.0 as the multiplier? Hard to argue with that one, right, Mr. Bernanke?]
Okun’s law also assumes there is some maximum potential GDP rate to generate the multiplier. This results in something like the NAIRU for the target unemployment rate. Using this in the law is a self imposed limit on growth, and another reason the ELR is superior.
We can use the TC rule with the ELR as a special case where (u-u) = 0 We are then left with a rule that targets inflation and population growth.
What is the proper inflation rate to use? Good question.
I used population growth as part of this equation because it is the largest single deflationary (or inflationary) pressure facing any economy. Maybe I am wrong, but I think of it like this:
100 agents, two groups of 50, with $50 total in the economy, no credit allowed, no desire for savings. Each day 50 of them produce 100 units of food, which is then of course priced at $1 per unit. The other 50 buy and then consume food and lay about. The groups switch every day in production and idle time. Then one day, 10 more identical agents show up with no money and equally join groups, so now each group has 55 in it. The price of the food produced must go down.
Deflation isn’t a problem in this toy economy, but in the real world deflation has much worse consequences than inflation. Anytime you can increase your wealth relative to other humans by sitting on your ass instead of working your ass off, there is a problem.
Anyway, thats all for now!