Monthly Archives: February 2019

The jeepney: decentralized trust in practice

21 Thursday Feb 2019

Posted by in Decentralization, EM

≈ 1 Comment

Have you ever ridden in a jeepney? Chances are that you haven’t, for even if you have visited the Philippines, unless you are a backpacker, you would have stuck to taxis. In our travels, we met expats who had lived there for years but had never been in a jeepney.

Well, you are missing on a practical lesson how decentralized trust works.

The jeepney is actually the most popular public means of transportation in the Philippines. It was originally made from the American jeeps left over from WWII. It is also the cheapest way to travel because of its open rear door design, the jeepney can pick up and drop off passengers anywhere. Having said that, it is not the safest way of transport, either mechanically (too old), or because of its seating configuration (a long bench with no seat belts, very low ceiling, combined with constant and sudden stopping).

What was fascinating for me was actually how its payment system works: it is all based on trust. Decentralized trust in fact. As it is not optimal to have a ticket collector, the jeepney driver is also tasked with collecting the money for the trip. The problem is that the entrance is all the way in the back. Not only the driver cannot collect the fare in advance, but he has to rely on other passengers to pass on the money to him. Many times he has to pass back change. The fare also depends on the destination which is shouted as the money changes hands, which adds an additional variable to keep track of. All this while driving, so obviously the driver cannot possibly follow up on all this!

But the system works. Passengers can see who has paid and who has not and they have an interest to keep its integrity in check not because of a fear of a fine (light regulation with minimal monitoring) but the realization that if it breaks, it means everyone has to take the more expensive and less convenient bus. Of course the system can also be ‘gamed’: if ‘majority’ of the passengers agree to cheat but the level/cost of cooperation is too high.

As societies become more complex it is basically suboptimal to rely on a centralized authority. An additional complication is if those societies do not have a strong institutional infrastructure, like in most of EM, or the traditional sources of authority start to be mistrusted, like in a lot of the developed world.  Decentralized trust then becomes a necessity and a prerequisite for maintaining the proper functioning of society. It seems to me that EM has the first mover advantage here and is likely to leapfrog the developed world.

Fiscal policy is next but it’s also unlikely to work

03 Sunday Feb 2019

Posted by in Monetary Policy, Politics, Uncategorized

≈ 1 Comment

Simon Wren-Lewis wrote an interesting article yesterday The Interest Rate Lower Bound Trap and the ideas that keep us there

Unfortunately, the ideas that keep us plugging pointlessly at monetary policy are not that dissimilar to the ideas which will push us into trying fiscal policy: both of them are based on using the old industrial model of labor and capital income distribution which is much less suitable in the digital age where technology takes center stage.

What particularly caught my attention was the 3rd paragraph and this very relevant question: “If these countries really did have a zero output gap, then why is inflation below target?” Which gets to the core of the issue about how technology has possibly substantially increased potential output.

Yet, our models do not fully capture that. Perhaps that is because we continue to put too much weight on capital and labor in the production function when clearly technology has marginalized them both, the evidence being in zero rates and flat wages.

Let’s take capital.

1) there is a large corporate capital surplus;

2) digital technology does not require so much capital;

3) consumer debt is maxed out.

All three of the above lead to low demand for credit meaning low interest rates regardless/independent of monetary policy.

So, after years of zero/negative/low rates (decades in Japan) it is finally obvious that the monetary transmission mechanism is now clogged (see above). Naturally, despite all the opposition, we are probably just a recession away to switching to fiscal policy.

But as labor’s turn comes, there is no guarantee and zero evidence (see, again, Japan) that fiscal policy would work as its transmission mechanism is probably also clogged. And the reason can be found in the fact that it is easier for corporates to switch from labor to technology in automating production.

A diversion.

That’s where the debate about technological unemployment comes in. And here I am in the camp believing that this time things are different because technology is more advanced and is taking away ‘IQ’ jobs in addition to just ‘brawn’. ‘EQ” jobs are humans’ last call of resistance but maybe not for too long.

Sure, no evidence of this for now but that’s because in the initial stages, with aggregate demand low, companies will choose to focus on cost reduction by using cheaper labor (taking advantage of the threat of automation keeping a lid on wages), than higher output/higher productivity using technology.

We’ve had jobless recoveries before but post GFC’08, we’ve had a ‘wageless’ recovery – plenty of jobs but anaemic wages. Neither is particularly good for aggregate demand as individual purchasing power barely increases.

The situation is even worse now as consumer debt to disposable income keeps rising (people now need two jobs to survive).

In the short run, we could potentially see a rise in wages as the labor pool gets gradually depleted, but the switch to automation would also be faster which would push unemployment up/wages back down. In the long run, technology substitution becomes inevitable as both its cost continues to decline and its capabilities to rise.

And, by the way, we are not helping, as apparently we are also getting dumber (see “Were the Victorians cleverer than us?” by M. Woodley et all).

Diversion ends.

So, the most obvious fiscal policy stimulus is infrastructure spending. That’s much easier to get voted in given the state of our roads and bridges, etc., and the fact that there are probably already too many people shuffling papers on desk jobs working for the government.

Infrastructure spending could be the most economically beneficial option but could also contribute the least to aggregate demand if it bypasses labor due to automation: awarding a billion $ contract to a company to renovate a bridge using mostly automated machinery is hardly going to increase labor’s purchasing power.

My feeling is fiscal policy will indeed soon become the default option. Sadly, not necessarily because it would work better overall for increasing aggregate demand but simply because it has become plain obvious that monetary policy is powerless.

Instead, we need to think ‘beyond the Overton Window’. The income transmission mechanism which we have adopted since the first industrial revolution, Work->Job->Income is broken. Monetary and fiscal policy thus become redundant. We need a new model more suitable for the digital age.