Saturday, March 11, 2017

Automation and Taxation

I am quite happy to have capital substitute for labor, especially when the labor is my own and the capital is inexpensive to purchase.  Examples abound.  I can't remember when I last did my taxes by hand rather than on the computer.  My dad did them by hand.  But that was 20 years ago.  Likewise, I can't remember the last time I contacted a travel agent to book a trip or the last time I hand wrote out a piece for some secretary to type. All of that seems like activities from a different era.

I am also not averse to using my (financial) capital to get somebody else to do the physical labor that I once did.  A few years ago we switched from me mowing the lawn to having a service do it.  Part of that is they do a more professional job, having tools to aerate the soil, putting down seed when that was needed, and in general keeping up with the yard maintenance in addition to the mowing.   (We live in a neighborhood where having a good lawn matters.) The other part is my arthritis and that sometimes the mowing got to be a burden rather than a form of relaxation.  For both reasons, this type of substitution makes sense to me.

On a personal level, this sort of capital for labor substitution should have two benefits.  One is to allow me to devote my (no longer so scarce) time to the usage I want to put it to, rather than to put in a lot of time into things that must be done but which provide no personal satisfaction in the doing.  (I count doing the taxes in this category.)  The other is to provide a potential health benefit when the labor that is done away with had some health risk attached to it.  Just to give a mild example here, using the dishwasher versus washing the dishes by hand, perhaps there is no health risk if you use rubber gloves when you wash the dishes.  But I don't like those gloves and tend not use them.  I suffer from eczema which the dish washing  mildly exacerbates.  The more sorts of labor that are somewhat detrimental health-wise that can be avoided, the better.

I lead off this piece with my personal views of this trade-off, because so many people are talking about automation and the elimination of jobs, as if that is pure evil.  What is different for the economy as a whole than in my own situation in this regard?  Let's focus on two points of difference.

The first is that I never experienced technology making my own efforts obsolete, quite the contrary actually.  In the mid 1990s the Internet enabled a new career for me in online learning.  The time was ripe for a change for me. (I turned 40 in 1995, our younger son was born the year before, and the economic theory I had been doing until then started to seem less compelling to me intellectually.)   It would be quite different now, especially if I were an adjunct not on the tenure track, teaching high enrollment introductory courses that might be outsourced as MOOCs or taught in some other way.  I would have me no job security whatsoever and might become alienated about doing the work now.  The lack of job security and alienation about work is a much more common occurrence now.  The second part of the 1990s is the last time I can recall a really tight labor market.  Real wages were rising then.  Opportunities seemed to abound.

The second is whether  other people have all these other things to devote their time to, things that they love to do but don't do enough of because they have to bring home the bacon first and foremost.  For me, writing is this type of thing.  It has never been a livelihood in itself for me.  But I have learned over the years that I need to have a mode of self-expression, a way to bring my internal reflections outside of myself, if for no other reason than that it enables me to move onto something else.  It hasn't always been this way for me.  Particularly before I started this blog, the job itself was the thing.  I found self-expression through work.   If people live for the job and the job is then taken away, they can lose their sense of purpose.

Let me note that while at the individual level each of these may seem like problems without a solution, at the societal level the first one definitely can be solved.  For example, many are now calling for a guaranteed minimum income.  That is one way to address the first issue.  I am going to leave the second issue without addressing it in this post.  I think I have a reasonable answer to it for people who, like me, were engaged in knowledge work.  I don't have an answer more broadly considered, so will let others ponder the question now and leave it to be taken up later.  It is now time to turn to the topic in the subject line of this post.

* * * * *

Yesterday or the day before I became aware that Bill Gates has called for the income that robots generate to be taxed.  If the robots are replacing workers, those workers previously earned income that was taxed. After this capital for labor substitution, that labor income is lost and with that the tax revenue that accompanies it.  Gates is calling for a taxation on the machines to make up for that lost tax revenue.  I found this piece, which describes Gates' ideas on the matter.

“Right now if a human worker does you know, $50,000 worth of work in a factory, that income is taxed. If a robot comes in to do the same thing, you’d think we would tax the robot at a similar level.”
Bill Gates, in an interview with Quartz

I found this suggestion more than a little amusing.  Indeed, one of the reasons I went through my personal experience at the beginning of this piece is to consider here non-factory work that has become obsolete as a consequence of technological improvement.  Steno-typists hardly exist anymore.  Word processing eliminated much of the function. Microsoft, obviously, played a big role in that.   Is Gates proposing to pay back taxes on all those lost earnings due to widespread adoption of Microsoft Office?

I assume he is not.  So I want to work through the economics of this a little more, talk about the economics some, and then try to identify where I agree with Gates and where I think what he says requires some modification.

First, it should be noted that while until this point in the piece I've talked about capital as a substitute for labor, instead it might be that capital is a complement of labor, meaning productivity of the labor is higher in quality or quantity after the capital has been installed.  A simple example is given by the online learning tools we now use in teaching, which we didn't have access to before the mid 1990s.  The effect of these tools is to expedite communication between students and their instructors, to allow instructors to become more aware of the work that students produce, because that work is readily accessible online, and to enable students to access much more content that might be relevant to them in learning the subject matter.  The Internet is a fantastic library.  These online tools have not done away with the need for good and devoted instructors (my earlier comment about MOOCs was about a brave new world that we really haven't reached and likely won't, particularly in regard to undergraduate education).  These tools have improved the experience for students and instructors alike.

As long as wages relate to productivity in a direct way, capital as complement to labor should raise wage income.  In this case there is actually increase in tax revenue from the installation of capital when it is a complement to labor.  So Gates' point is not relevant here.

Second, there is the issue of whether the IRS or state taxing authority can distinguish the substitutes case from the complements case.  If not, so that a general tax increase on capital would be put into place, let us note that this would create some bias against the use of capital when it is complementary to labor.

Third, one should recall the example of Steve Jobs, Bill Gates' friend and rival, who famously took $1 as his pay for being CEO of Apple Computers.  He took the rest of his earnings as stock and stock options. This was partly marketing ploy.  But there was a real reason for doing this as well - tax avoidance.  The dividends and capital gains Jobs received were taxed at a lower rate.  Earned income rates, which is what determine what most of us pay in income taxes, are higher than unearned income rates.  Further, there is tax avoidance with capital income as financial capital is highly mobile and readily can be moved to low (no) tax jurisdictions.

Fourth, capital's share of GDP has been rising (so labor's share has been falling).  There are multiple possible explanations for this.  I prefer the market power explanation.  Labor's market power has fallen while management's market power has risen.

If you couple the third and fourth reasons, you get that tax revenue as a share of GDP has been declining and will continue to decline further in the future.  This is happening without any adjustment in tax rates.  This is where I agree with Gates.  It is a distinct issue, one that must be addressed.

I like to push arguments to logical extremes to see what that does to things.  Consider a world where there is no paid work at all.  Machines do it all.   Individuals do receive a guaranteed minimum income.  They also still consume government provided goods and services.  They drive on roads the government pays for.  They travel through airports that are likewise publicly provided.  Robots now do all the police work.  Other robots teach our children.  Water and sewage systems also are publicly provided and get regular capital upgrades.  And on and on.  All the goods and services provided by government must be paid for out of capital income.  In this world there has to be balance between tax revenue generated out of that capital income and government expenditure on the goods and services that people need and want.

How do we get there from here?

No comments: