Saturday, April 21, 2018

Welcome the robots.

There is a never ending debate about whether automation will eventually create problematic unemployment.

The same fears have been around since the dawn of the industrial revolution.  If you wanted to suggest to a farmer in 1840 that he and almost everyone he knew would become obsolete, and it would be overwhelmingly for the better, he would have been unpersuaded, if not downright angry.

Likely, he would have hastily demanded an answer to the question, "What in the world would you have us all do to earn a living?"

You could, quite reasonably, have answered, "Well, maybe instead of farming, people could do air ballet for strangers, and, I don't know, sell tickets or something."  Or, just as reasonably, you could have described most of the jobs at Google, or Salesforce.com, or many other firms in today's economy.



The farmer would, understandably be non-plussed.  It's not just that you would have a hard time explaining what you meant to him.  It's that there would be no way to get him from point A to point B.  The explanation would have simply been beyond his imagination.

This is what growth means.  It means that things we take for granted are replaced with things we can't imagine.  Luckily, we can imagine air dancing today.  Automation and healthy economic growth will bring new things we can't imagine.  I would like to eventually take for granted the things that today I can't imagine.

It is true that there are jobs because there are needs to be met.  It is also true that there are jobs because there are people looking for needs to meet.  Unemployment is a function of frictions over limited time periods, not of a lack of demands.  This is so clear, it is strange that we have to remind ourselves of it.  Certainly the places where the most broad supply of services are available are not the places where there are more people lacking in available productive activity.  The opposite is clearly more true.  There was a time when this person, looking for needs to meet, would have mostly been looking at a potential set of needs that required leading oxen through a field or some such activity.  Today, that set is large enough to include air dancing.

Should we want change and growth?  To answer yes requires that we cease to rely on our imaginations.  It is an explanation destined to lose, and so we will likely be relitigating it, still, a century from now, when air dancers ask, indignantly, "Well, then, if you invent better holographic robot dancers, what in the world are we supposed to do?"

10 comments:

  1. > Unemployment is a function of frictions over limited time periods, not of a lack of demands.

    That's when the regulation doesn't get in the way even more.

    In East Germany in the 90s in addition we had basically West German levels of welfare, but ex-socialist levels of productivity.

    Unemployment was high, and it took until at least the mid 2000s to even start to budge. (It's still twice as high at about 10% in the East, as the about 5% in the west. But that's a change from the 90s numbers of about 20% and about 10% respectively.)

    I think eg South Africa also peraistent high unemployment.

    Of course, none of this distracts from your argument.

    What does distract from your argument that anything that happens has to happen for a first time some time.

    Eg what when the industrial revolution kicked in (or perhaps a bit earlier with the agricultural revolution) things did change: we started escaping the Malthusian constraints and living standards rose.

    AI could be another moment like that, but for human labour demand?

    (Mine is not a positive argument that this is indeed the case. If I had to bet I would actually support your version of things. I am pointing out that your argument here is perhaps not strong enough to carry your conclusion.)

    ReplyDelete
  2. OT but maybe worth a post.

    There is a famous libertarian Richard Epstein (Hoover) who has posited that when communities zone properties, but the zoning decreases the property's value, then the owner is entitled to compensation.

    An administrative nightmare, but makes sense.

    But…think about it. How about when (somewhat self-imposed) zoning creates artificial scarcity, enhancing value?

    Should not such property owners pay a tax?

    ReplyDelete
    Replies
    1. Land Value Tax solves this just nicely without requiring anyone in the government to decide what impact the taxes have.

      If unimproved land prices go up for any reason: more LVT to be paid. If unimproved land prices go down for any reason: less LVT to be paid.

      Doesn't matter whether that impact is from the area getting randomly fashionable, or changes in regulation, or a new train station.

      Delete
    2. Should have been:

      > Land Value Tax solves this just nicely without requiring anyone in the government to decide what impact the zoning changes have.

      Delete
  3. Matthias:

    Upon reflection, I think you are right.

    A national property tax at 5% of value might be the ticket.

    There is about $30 trillion of housing in the US. I think commercial somewhat less,

    So say $40 trillion. A 5% national annual tax on property, along with some fossil fuel taxes, and we could just about eliminate income taxes, or Social Security taxes.

    ReplyDelete
  4. Benjamin, that's roughly the calculation to do.

    There are some interesting interactions, of course.

    First, you actually want to do your calculations based on rent, not capitalized rent (ie prices). That's because interest rates and the very LVT we are talking about will impact the prices.

    Second, getting rid of most taxes on labour and capital increases land rents, thus an LVT might make lowering other taxes pay for itself.

    (Similar with infrastructure improvements: sensible public infrastructure improvements will raise land rents by more than they cost, thus paying for themselves. Right now, publicly financed infrastructure improvements are often captured by private land holders.)

    This is all pretty standard Georgist fare.

    The other side of the coin of lowering taxes on labour and capital is paying out a basic income. A basic income would also raise land rents. Thus without an LVT it would mostly flow to land owners, and only a part of it would actually enable people to increase their consumption.

    With an LVT, the part of a basic income that goes to increase land rents will just get recycled back into the tax base.

    See https://medium.com/basic-income/why-land-value-tax-and-universal-basic-income-need-each-other-42ba999f7322 for someone doing the numbers on that.

    ReplyDelete
  5. Matthias:

    Well, great ideas.

    For years I have said the US should migrate away from taxes on income and labor, but….

    I suspect that by now, people who are influential know how to shield income from taxes, though law, regulation or offshore bank accounts, so there will be no tax reform.

    If we went to property taxes, there would be no dodging taxes…so that idea is stillborn.

    Still, in a dream, we would phase in property and fossil fuel taxes, and phase out income and Social Security taxes…say over a 10 to 20 year period….

    I also like import taxes, as it is pretty hard to sneak cargo containers into the country….

    BTW, Singapore says it is a no duty country, wide open and free...oh, but they have a 7% tax in imports….

    Well, we can dream….

    ReplyDelete
  6. I wish some import duties were lower.

    ReplyDelete
  7. Benjamin, it's a bad idea to tax transactions. Gains from trade and all that. Causes lots of deadweight loss. So don't tax shipping containers.

    I happen to live in Singapore (again since December this time). Great place. It's the promised land. A weird and pragmatic mixture of libertarian's wet dream and some aspects of nanny state.

    ReplyDelete