Tuesday, June 17, 2014

After Reading Piketty’s CAPITAL


Istvan Gorog   
June 2014, Eureka, CA

I posted my “Notes on the Economy: Jobs, Incomes, and What the Future May Hold” contemporaneously with the American publication of Thomas Piketty’s “Capital in the Twenty-First Century”. Here I comment on our similarities and differences, as well as append my earlier thoughts based on what I learned from reading Piketty.


Contents

  • Introduction
  • Overview of similarities and differences
  • Education and unemployment
  • Public debt
  • Intellectual capital
  • More on inequality
  • Conclusion
  • Notes


Introduction

In April of this year I posted my “Notes on the Economy: Jobs, Incomes, and What the Future May Hold”.  Also in April Harvard University Press released the English version of Thomas Piketty’s “Capital in the Twenty-First Century”.  Piketty’s book is a 700 page major scholastic achievement, a detailed study of wealth and income inequality, based on the detailed analysis of historical economic records. My Notes are a brief summary of my observations supported by easily available data. Nevertheless our overlap is sufficiently strong to compel me to write this comparison of my modest work and his bestselling masterpiece. Principally, my interest here is to summarize where his data supports my findings, where my Notes supplement his work, and where our emphases differ. Briefly: we do not disagree, we simply focus on different aspects of the same problem, and offer complementary solutions that together will be needed to save American capitalism and establish a just society in the 21st Century.

I believe it is important to emphasize that we both believe that capitalistic democratic society is the best social system known to date. While as I wrote in my earlier Notes,  some may perceive some of my views as socialistic, and others have charged that Piketty is a Marxist, we both seek to repair the growing flaws of capitalism to avoid its demise through social discord between the haves and the have-nots that ultimately could lead to destructive violence.


Overview of similarities and differences

We both recognize a significant inequality where the upper brackets’ shares both in terms of wealth and income are at historical highs and are still growing. Piketty addresses this issue globally, focuses on its evolution over history. It is this growing inequality that he believes needs remedy. I look only at the USA, using some data from abroad only as guiding reference. I focus principally on the growing unemployment resulting from technological advances. In my view these advances now more and more substitute robotic machinery, thus capital, for labor. Furthermore, I argue that virtually no job or profession is immune from this substitution of human activity by intelligent machines that can move, are dexterous, can see and may have also other sensory capabilities, obey voice command, and are easily programmable. They will be mass produced, affordable, and ubiquitous. I foresee them in factories, offices, homes, and public places; they make things, answer questions, deliver goods, and provide personal assistance. I believe that we can produce enough food, goods, shelter, transportation, and energy from renewable sources to provide for all our needs, but with our economy as it is structured today, in the 21st Century we will not be able to provide fulltime jobs for all to earn enough to pay for all their needs. There simply is not sufficient need for labor in our economy; thus there are not enough jobs today, and in the future there will be still fewer. The solution to this quandary is Piketty’s social state with more progressive taxes on income than we have today and a new general wealth tax on wealth.

I also believe that it is better for all to be employed than for some to work and others to be unemployed; therefore the legal working hours need to be adjusted with the aim to achieve full employment with reduced working hours. Also, contrary to much of the current political wisdom, for the same reason the retirement age may need to be reduced rather than increased. Piketty’s social state, paid for by progressive taxes, could address these needs.


Education and unemployment

Even though education is not an absolute protection against unemployment, its significance in employability cannot be overemphasized. As I discuss it in my Notes, there is a direct correlation between attained education level and employment. The “official” unemployment rate (US BLS U-3 measure) in 2013 for people with less than high school diploma level was 11%, for those with high school diploma was 7.5%, while for people with Bachelor’s degree or higher level was 3.7%. The “total” unemployment rate (US BLS U-6 measure that includes all who wish to work full time but can’t find such employment and either work part time or have given up looking for work) is about twice that of the official rate: 12.3% vs. 6.3% in April 2014 for the USA. (No BLS U-6 data is available by educational segmentation.) Thus assuming a constant total-to-official unemployment rate ratio for all education levels, over 20% of the population without high school diplomas  and over 15% of those with high school diplomas but without additional education is unable to find fulltime employment. The foregoing is likely to be an under-estimate; less educated social groups are more likely to be discouraged and give up looking for satisfactory employment than are the better educated, more successful groups. Furthermore, based on personal observations I estimate that in fact the majority of the younger people (under about 35 years) with only high school diploma or less education are unemployed or underemployed. More than 40% of the US population has high school only or less education; thus their employability status is a major social and political issue. Piketty views access to education everybody’s need and right in an equitable society and thus the state must provide it at public expense, similar to how it is done in much of continental Europe.


Public debt

Public debt is a major concern of many Americans and it is much discussed in the media; frequently there appears to be some confusion about who owes what to whom. Neither Piketty nor I consider it a top priority. Before discussing it, I wish to clarify certain aspects of it. First I wish to emphasize that there are two kinds of national debt that are not always clearly differentiated and thus are frequent sources of confusion and misconception in public discussions. There is public debt and there is external debt. Some of the public debt is external debt and some of the external debt is public debt, but the two are not the same. The public debt, also known as sovereign debt, is what the government owes, what it borrowed to do what it is supposed to do: provide for the defense, education, health, and welfare of the people, as specified by the laws and provided for in the budgets; both the budgets and laws have prior approval by Congress. The external debt is what the US government, institutions, businesses, and individuals collectively borrowed abroad and thus owe to foreign governments, institutions, businesses, and individuals. Thus in principle, we could have no public debt and still have foreign debt, or no foreign debt and still have public debt; in practice we have both.

The US public debt on May 31, 2014 was $17.5 Trillion, out of which $5 Trillion was intragovernmental, whereby the government borrows for its current operational needs funds that it currently collects, but will need to pay out only later. The principal source of intragovernmental borrowing is Social Security. The public debt net of intragovernmental borrowing, thus the debt held by the public, is $12.5 Trillion, or 74% of the GDP. (In 2013 the US GDP was $17.1 Trillion). However, the debt held by The Federal Reserve is included in this figure of public debt. The Fed’s share of this is $2.4 Trillion. Thus netting out the Fed holding, the remaining debt held by the public is $10.1 Trillion, or 59% of the GDP.  (To summarize the confusing terminology: we have public debt that includes intragovernmental debt; the debt held by the public is the public debt net of intragovernmental borrowing; the debt held by the public includes the debt owed to The Federal Reserve; the debt held by the general public – which may include foreigners and foreign governments – is debt held by the public net of The Fed holdings.)

The external debt of the USA as of December 31, 2013 was $16.5 Trillion, of which $5.9 Trillion, or 36% of the GDP, was government debt. The US also has external financial assets and the difference between the external debt and the external financial assets is the net international investment position (NIP); at the end of 2013 the NIP of the USA was negative $4.6 Trillion, i.e. foreigners owned American assets in excess of foreign assets owned by Americans worth about 27% of the GDP. According to Piketty, the total private and public capital in the US is about 450% of the GDP, thus the US NIP is about 6% of the total US capital.
It is also important to recognize that $13.5 Trillion out of the total US external debt of $16.5 Trillion, or over 80%, was borrowed in US dollars. Thus in principle the Federal Reserve could wipe out much of the external debt by simply printing more dollars. Clearly there would be significant damage to the US prestige and economy as a consequence of such an action; nevertheless it is an option.

To put the above numbers into a Global perspective, the global external debt (the sum of all external debts of all countries at the end of 2013 was $73 Trillion while the GWP (Global World Product, i.e. the sum of the GDPs of all countries) was $74 Trillion at official exchange rates and $87 Trillion at purchasing power parity (PPP). Thus the US external debt as percent of the GDP is in line with the global figure. Obviously the global NIP, the sum of the NIPs of all countries, is zero. Since the US NIP is negative, we conclude that foreigners feel more secure owning American assets than do Americans owning foreign assets.

Having established the basic quantitative parameters of or national debt, I wish to address some key conceptual issues related to public debt. Here I will follow Piketty’s reasoning that I adopt henceforth as also mine. When the taxes collected are insufficient to pay for Government expenditures, public debt is accrued as the Government finances its expenditures by borrowing. In other words, one way of looking at public debt is that it a substitute for taxes. Another way is to view it as regressive taxation: the rich collect rent (interest) from the Government by lending to it and all tax payers share in paying for this rent (and one way or another, ultimately also will need to pay back the principal). – The third view, much cited in certain circles is that it is a way to pass on the cost of current generation’s profligacy to future generations, is not very credible, since in the past two centuries this never happened, even though government debts in the USA, and in all developed countries, continuously existed during this period, at times at significantly higher levels relative to GDP than is the case now in the USA. -- Public debts were managed (old outstanding ones virtually eliminated) by growth of the economy (USA through its history), inflation (most everywhere in the 20th century), and forgiveness (post WWII Germany).

Historically the wealthy like public debt as means to derive a relatively safe income from accumulated capital. They do not like inflation since it reduces the real value of the rent collected, as well as of the principal. By the same token, the wealthy like deflation since it increases the real value of the interest collected, as well as of the principal. Today most economists agree that deflation is dangerous; it stifles growth and leads to recession. The generally held belief by experts both within and outside of government is that a moderate inflation of about 2% per year is desirable.

So how are we going to manage our debt?  Can we outgrow it? Not likely. Historically, in addition to productivity growth, the growth of the US economy was aided by territorial and population growth. Going forward GDP growth will only come from productivity and population growths.  Let’s use a specific time frame: what can we expect in 36 years, by 2050? As Piketty emphasizes, the real inflation adjusted GDP per capita growth historically in America over any extended period has never exceeded 2%, and 1.5% maybe a realistic forecast (see Table 2.5 in Piketty’s “Capital…”). According to the US Census forecast, by 2050 the US population will be about 400 million vs. 320 million now; thus the population growth rate is forecast to be about 0.6% per year. The combined effects of per capita productivity growth and population growth are expected to result in a real GDP growth by 2050 of about a factor of 2, i.e. we can expect the GDP of 2050 to be about $34 Trillion in 2014 dollars. Thus today’s publicly held debt of $10.1 Trillion, net of the Fed holdings, will still be 30% of the GDP; the total public debt of today, including the Fed holdings and intragovernmental borrowing, would in 2050 still amount to 50% of the GDP. Since we cannot outgrow it, the remaining options to manage our public debt are taxation, inflation, and default. Outright default is extremely unlikely. Inflation, especially one induced by the Fed, is a form of default; again it is an unlikely US debt reduction measure. Thus the likely solution is increased taxation. Piketty advocates an increase of the top marginal income tax rate to 80% and the introduction of a wealth tax of a few percent per year. While the political will to adopt these measures in America is lacking today, increased taxation is a necessity and the wealthy are not likely to escape paying their share. Whether increased taxation will come according to the formula Piketty suggests, or in some other manner, is not clear to me. But I am certain that it will come, because it has to come.

Currently we have public debt because as a country we needed to pay for things that were needed, and we were unwilling (politically unable) to raise our taxes, and there were (are) people willing to lend us. Our public need for public funds will only increase and not decrease.


Intellectual capital

I see a world rapidly emerging where smart machines do most of the work and much of the need for traditional labor disappears. To install and operate the smart machines, society needs two kinds of capital: physical and intellectual capital. Here I would argue that the traditional term “human capital” for the acquired non-tradeable but rentable human skills is not adequately clear in the 21st Century. All of us people are born human, but going forward our sheer human competence has no economic value; it is our learned skills that society needs for its growth and operation. Therefore I prefer the term intellectual capital over human capital.  Owners of capital collect “rent” on the capital deployed. The rent taken by owners of physical capital is the interest and royalties earned; the rent taken on intellectual capital is in the form of wages and professional fees earned (Note 1). Physical capital is transferrable (sold, given away, inherited, and even stolen), intellectual capital is not transferrable. Thus intellectual property (IP), be it patents, copyrights, or designs, is in this sense physical capital. It is interesting to note that while intellectual capital is non-tradable and cannot be inherited, there is increasing evidence that social mobility does not increase with the increasing emphasis on intellectual capital. It appears that individuals born into families who possess physical and/or intellectual capital will more likely acquire intellectual capital than do less fortunately born individuals. To acquire physical capital, an individual may inherit it. To acquire intellectual capital, an individual needs training, which in turn also requires both physical and intellectual capital investments: physical to cover the living expenses and intellectual to impart the knowledge. The fortunate have access to both early on in the homes in which they grow up. The significance of access to intellectual capital during childhood may be greater than in later life. Intellectual capital appears to be similar to physical capital in a sense that a seed investment is needed so that that when it is well managed it may grow significantly. Without a starting seed no amount of compound interest can lead to wealth. Similarly without a starting interest in knowledge acquired in childhood, no matter how much access to education may be available, no significant level of intellectual capital can one acquire. It is of further interest to recognize that more and more advanced  teaching is taking place via remotely accessible electronic means, whereby one highly trained individual may impart his/her knowledge to thousands or even millions of  students. Thus in advanced training the relative need for physical capital over intellectual capital is increasing.

In any case, recognizing the significance of intellectual capital and the growing insignificance of traditional unskilled labor, Piketty’s social state will need to pay increasingly for education and welfare. The money to pay for these increasing expenditures will need to come from those who possess capital, be it physical or intellectual.

Here I wish to separate the unjustified supersalaries from the justifiable compensations derived from intellectual capital. Earnings derived from intellectual capital put many professional in the upper decile of total income; some even make it into the upper centile, earning hundreds of thousands of dollars per year). The supersalaries of top tier corporate managers (possibly  millions, or even tens of millions per year, are not justifiable by their contribution to the economy, not by their competitive value. As Piketty argues, supersalaries are the result of  crony behavior on self-serving corporate compensation committees. Thus, while both intellectual capital and supersalaries contribute to inequality, earnings derived from intellectual capital are the result of useful contributions to the modern economy; supersalaries are derived from cronyism.

Physical capital has two major principal components: real estate and financial assets. As cited above, that the total US wealth, i.e. physical capital, is about 4.5 times the GDP, thus it is $77 Trillion. At 5% return on investment, US wealth earns about $4 Trillion per year. I am about also interested in estimating the magnitude and associated earnings of intellectual capital possessed by professionals in the USA. For the purposes of my estimate, in this group I included engineers, scientists, physicians, lawyers, architects, and accountants. I estimate that there are a total of about 9 million people in this group, with about two-thirds in the engineers and scientist category (Note 2). Then, using a published figure of $1.1 million for the average intellectual capital value owned by professionals (Note 3), I estimate the total US intellectual capital as $10 Trillion. Owners of intellectual capital also tend to accrue physical capital by systematically saving and investing from current earnings, such savings may take the form purchasing a home and/or investing in financial assets for retirement purposes. As the result of such savings and accruals, the total physical capital owned by professionals is likely at least to match, but even more likely to exceed the total intellectual capital owned by them (Note 4).


More on inequality

Now I wish to return to the question of inequality. Everybody who looks at the widely available statistics agrees that we have significant inequality. In my “Notes on the Economy…” my basic thesis is that 1) a new type of fundamental inequality is of growing significance: technological change forces more and more people out of the productive and growing economy (Note 5); and that therefore 2) in a democratic just society the fortunate, whose share in the benefits produced by this economy is growing, must take care of those of those who became marginalized. Furthermore I argue that as machines replace people, the role of capital inevitably increases. Piketty provides detailed analysis of historical data and he concludes that when the rate of return on capital exceeds the growth rate of the economy, capital’s share in the GDP increases and to rebalance the resulting inequality appropriate tax policies need to be adopted.  Furthermore he shows that throughout history capitalism has not been simply operating according to some divine law of the free markets, but was much affected and controlled by wars, depressions, and political acts. He argues that inequality is now dangerously growing and he places (at least partially) the blame for the 2008 Financial Crisis, and for the Great Recession in 2008-9 that followed it, on this inequality.

The recovery from the Great recession is not proceeding as many would wish and expect. In June 2014, about five years after the recession, we have 2.1% per year GDP growth and the US unemployment is at 6.3%, while at the start of the recession GDP growth was about 3% and unemployment was 5.0% (Note 6). To make the unemployment number worse, we need to recognize that the labor force participation  rate dropped from 66.4% in January 2007 to 62.8% in May 2004 (Note 7). Thus, even though now fewer people are looking for jobs than before the Great Recession, still fewer can find one. As I cited earlier, according to Piketty, the GDP per capita growth rate over any extended period never exceeded 2%, and going forward it will more likely be about 1.5%. Since now the US population is growing at about 0.6% per year, practically zero, the real GDP growth must come from productivity growth. Thus 2.1% GDP growth rate, especially with increased unemployment, is rather good by historical standards. This may indicate that the economy is in fact still catching up for ground lost during the recession, or that in fact we may have already fully recovered from the great recession, but we do not like what a recovery in the new economy looks like.

Decreasing employment participation and increasing unemployment rates are well aligned with my thesis that jobs are disappearing permanently due to technology changes; or stating the obvious more explicitly in a different way: in the new economy technology advances eliminate more jobs than new opportunities resulting from new technologies create new jobs. We must face up to this fact and act accordingly. For the benefit of all, the new societal need for sharing must be recognized and new rules to establish and maintain the needed sharing must be legislated and implemented.


Conclusion

Several conservative and libertarian attacks on Piketty’s work have been published, notably in The Economist, The Financial Times, and The Wall Street Journal. Some claim that there is no growing inequality; others call him the new Marxist. I disagree with these attacks. Paul Krugman twelve years ago raised the issue of increasing inequality, and he recently pointed out that some of these attacks are “politically motivated efforts to deny the obvious” and also are comparing apples and oranges in order to discredit Piketty’s work (Note 8). It is notable that earlier bestselling books (Note 9), one by Paul Krugman and another by Joseph Stiglitz, both Nobel laureate American economists, addressing the same issue of inequality, both books shorter and less filled with quantitative data and thus more readable by the general public, did not receive the same level of public attention as did Piketty’s. Maybe the times are changing?

There is one fundamental issue on which I disagree with Piketty and the other above cited economists. They all emphasize political action to stimulate the economy, to assist people to move from jobs being lost to jobs being created. They seem to feel that the balance of the job market as we knew it in the 20th Century can be reestablished with appropriate political actions. I disagree. As I stated above, I believe that we are faced with a fundamental structural change, where inevitably in the 21st Century technology will eliminate more jobs than create new ones. We must develop new social and political thinking to address this fundamental issue (Note 10).

One may indeed ask whether Piketty’s ideas, and mine too, are in fact socialistic. Here again I wish to state most emphatically that I am an ardent believer in capitalism and I believe so is Piketty. I strongly believe that supply and demand must be mitigated by a competitive market. In the 20th Century we witnessed the failures and miseries of centrally planned economies, most notably in Stalin’s Soviet Union and in Mao’s China. At the same time the competitive market must be subject to laws and regulations. Too big to fail is unacceptable: why should society pay for the blunders of excessively compensated superrich managers because their organizations cannot be allowed to collapse? This is clearly wrong. Also environmental issues, renewable energy, and public safety need to be addressed by appropriate regulations. Furthermore, taxation should be used to provide equitably for all. Those who harvest the benefits of the modern economy must provide for those who in fact are being rejected by the same forces that allow the fortunate to harvest the benefits. Inequality must be reduced and democratic control over capital reestablished.




Notes

Note 1 Future returns of capital can be computed, or estimated based on past experience, in various ways. For physical capital the simplest method is to use is to calculate the compound interest over a fixed period, more sophisticated methods may use simulation calculations that employ random sampling of historical data. For returns on intellectual capital, various methods have also been developed based on the investment, i.e. the cost of education, and the anticipated excess earnings attributable to the intellectual capital acquired through the education. According to such a calculation, a UC Berkeley engineering degree in 2013 costs, after financial aid, about $70,000 and the graduate engineer over 20 years will have earned $1.1 million more than a cohort who never went to college (The Economist April 5th 2014).Thus one can assign a dollar value of $1.1 million as an estimate for the average intellectual capital value owned by the graduate engineer.  For comparison, at 5% compounded interest, after 20 years a $70,000 investment will have earned about $0.1 million.
Note 2 The total engineering and scientific manpower estimate is given by The Congressional Research Service as 6.2 million people (CRS report R43061, February 19, 2014). For other professional manpower estimates, see United States Department of Labor, Bureau of Labor Statistics, Occupational Outlook Handbook (http://www.bls.gov/ooh/home.htm). 
Note 3 Combining the data sources cited in Note 2 above, physicians represent less than 10% of all US professionals. The average cost of education and average earnings of all other professions is about the same, thus I use the value of $1.1 million from Note 1 as a crude estimate for the dollar value of the average intellectual capital per professional.
Note 4 If one saves $10,000 every year, an amount most mid-career professionals can afford, and invests it at a real net of inflation rate of return of 3%, after 20 years the resulting physical capital value is $2.6 million; then 9 million professionals will own physical capital of $23 Trillion total value, which is 30% of the total US wealth of $77 Trillion. – This is likely a significant overestimate, since even in the most oversimplified model of such a group not everyone in the group can be expected to have saved uniformly  for 20 years. If we assume that the number of group members is constant over the years, that for all participants the group participation ends after 20 years of participation, and that every year the same constant number joined (i.e. membership is evenly distributed over the years of participation), then group average accrual would be about $1.2 million per person, and the group’s total physical capital accrual would be about $10 trillion or 13% of the total US wealth.
Note 5 In my Earlier “Notes on the Economy…” I focused first on manufacturing automation and then discussed how not only manufacturing, but also most service activities are likely to be taken over by smart machines. Lest one thinks that artists are safe from losing their jobs to machines, now opera orchestras may face replacement with “the digital sound of sampled instruments”, see “A Digital Orchestra for Opera?”, NY Times, Thursday, June 12, 2014 p.A1.
Note 6 See: “A Scarred Economy”, NY Times Thursday, June 12, 2014, p. B1. For GDP growth data, see: US Department of Labor Bureau of Labor Statistics, National Income and Product Accounts Tables, Table 1.1.1. Percent Change From Preceding Period in Real Gross Domestic Product. For the three year period preceding the Great Recession, years 2004-2006, averaging the Bureau of Labor Statistics quarterly data indicates about 3% real GDP growth rate per year.
Note 7 US Department of Labor Bureau of Labor Statistics, Economic Releases, Databases, Tables & Calculators by Subject (http://data.bls.gov/timeseries/LNS11300000)
Note 8 Paul Krugman: “On Inequality Denial”, NY Times, Monday, June 2, 2014, p. A17.
Note 9 Paul Krugman: “The Conscience of a Liberal”2007; Joseph E. Stiglitz: “The Price of Inequality”2012
Note 10 Some skills and professions will continue to be in demand, offered by owners of intellectual capital; at the same time the number of unemployables, those without intellectual capital, will inevitably grow. -- In my view the primary problem with the superrich is the superpower they wield in setting the political agenda. The problem of inequality is more broadly based: on the fortunate side are the owners of intellectual and physical capital; onto the unfortunate side go more and more unemployables; and the middle is progressively disappearing. In my view not only do we need to reduce excessive concentration of wealth and power at the very top, we also need to increase more broadly the sharing in the benefits of our productive economy across society with all who become structurally marginalized. Who will have to pay for it, i.e. to share their benefits? According to a recent CBO report on incomes (http://www.cbo.gov/sites/default/files/cbofiles/attachments/10-25-HouseholdIncome.pdf), the top 1%  earners take 21 % and the top 20% earners take 55% of all income.  (70% of the one-percent-group are managers and professionals from medicine, law, and finance; a group whose earnings clearly are derived from their intellectual capital.) Wealth is even more concentrated at the top than is income: the wealthiest 0.1% owns more than 20% and the upper 10% owns about 75% of the total wealth of the nation (E. Saez and G. Zucman, 2014 slide presentation, http://gabriel-zucman.eu/files/SaezZucman2014Slides.pdf). --  Increasing welfare rolls are inevitable, but every effort should be made to combine welfare  subsidies and new employment opportunities. To this end we need to reduce working ours, make education at all levels accessible to all, and make public investments to provide incentives to create new jobs where the new economy on its own shows at best limited growth. (Note that Google’s 2014 financial estimates forecast $40 Billion revenue and 50 thousand employees; this indicates excellent productivity of $800k per skilled employee, but does not describe an engine for broadly based job creation. In May 2014 there were 135 million employed Americans [http://www.bls.gov/news.release/empsit.t05.htm] who will produce  a GDP of $17 Trillion, or $126k per employed person.) --The principal areas to address with new public incentive investments are: renewable energy, infrastructure, environmental cleanup and protection. Maybe a new new-economy can be thus created, but even then the employment prospects of the unskilled and skill-mismatched are dim and their support by the social state will be required.

 



Sunday, April 27, 2014

Notes on the Economy: Jobs, Incomes, and What the Future May Hold.


Istvan Gorog   
April 2014,  Eureka, CA

Deficit, inequality, technological progress, and unemployment are interrelated. Here I argue that a major rethinking of our economic and social policies is needed to preserve our American capitalistic society.


Contents

  • General background thoughts on deficit, inflation, and growth
  • Automation and employment
  • Where we are coming from and are going to
  • Concluding remarks: a Global perspective
  • Appendix: Quantification of American income and wealth trends.



General background thoughts on deficit, inflation, and growth

In the USA we have widely discussed concerns regarding the economy, especially the deficit. I share such concerns in general though my primary focus is elsewhere. Yes, the national debt is high, the second highest in US history as % of GDP. The highest (~120%) was reached after recovery from the Great Depression and ending WWII. Now we have about 80% as we are recovering from the Great Recession and ending two of the longest wars in US history. (And hopefully avoiding a new one with Mr. Putin.) While the debt is still increasing, the deficit is now decreasing (now ~3% of GDP vs. ~10% a few years ago) and is expected to do so for several years before it starts to rise again after 2017. Currently and in the near future, the debt is not a problem because interest rates are now very low.

Why are the interest rates so low? I think the answer is twofold, neither one very encouraging. First, mostly cited in the media is that the world economy and politics are so uncertain that people with money consider US Treasuries the preferred safest investment, even if the payback is nil. The second part of the answer must be lack of demand relative to available supply, potential GDP exceeds actual GDP; more is offered than is taken. -- High-interest rates reflect that people want more stuff than what is available for them to buy; this is how rampant inflation developed in some European countries after WWI and WWII. (Even though most people are not well off after a war when both industry and agriculture are destroyed, people are desperate to spend whatever they may have on essentials they desperately need.) Now we have no inflation, and some even fear deflation, considered a nightmare scenario in modern economics.  Why is there a lack of demand relative to supply? Because there is unemployment, which also results in stagnant wages.

Why is the deficit expected to grow again after 2017? Because the economic growth is expected to slow, because by then supply and demand are expected to be in balance (potential GDP equals actual GDP). GDP growth will slow partly due to an aging population. So what is going to happen to the national debt? If unlike after WWII, we cannot outgrow it, we will have to pay it back. There is only one way of paying it back: to pay it back, taxes must be increased. There is much talk about reducing entitlements and government waste. Well, our entitlements are small relative to other advanced economies, and we need more infrastructure and green energy investments.  Also, to preserve American democracy and social peace, we need to re-balance income distribution and reduce the growing difference between the average and the median household incomes. For several decades now, average inflation-adjusted income grew, while the corresponding median income stagnated or even declined. All the GDP growth since the 1970s went to the upper crust (this includes the lucky well educated), while the largest segments of society saw no improvement in their lives. This situation is unjust, and unsustainable, leading to resentment, class warfare, and political instability. It must be reversed. To accomplish this reversal, we need to return to steeply progressive taxation without loopholes. Yes, Warren Buffet and Mitt Romney should pay not a lower, but a much higher percentage of their incomes in tax than their secretaries do. Then we will be able to maintain a stable democratic society and pay back the debt even in the absence of sufficient GDP growth to repeat the historic precedent.


In the worst-case, nations like the USA, unlike ordinary households and unlike Greece in recent times, have another way to wipe out the national debt. They can print their own currency as necessary and there is not much anyone can do to stop that. Of course, there would be a price to be paid: loss of credit, loss of international trade, loss of general national trust, most likely followed by an extended period of a declining economy, decreasing GDP, and hardship for all. Nevertheless, it is an option available as a last resort to balance the nation's books.


Automation and employment

Now I want to focus on what I consider to be the primary long-term economic issue: employment/unemployment. Let’s start by considering what industries drove economic growth and where were the job creators. Research indicates that in the recent past economic growth came from the tradeable sector, while jobs were created in the untradeable sector. Here tradeable refers to those economic activities that can be performed elsewhere and the results of the activity imported; e.g. most manufacturing, and agriculture. The untradeable (or not-tradeable) economic activities are essentially local, they must be performed where the demand is located; e.g. health and educational services, and construction. While the boundary between tradeables and untradeables is not sharply defined nor fixed forever (e.g. recently part of the rebuilt Eastern span of the San Francisco Bay Bridge was built in China, and online education can be imported), the concept is clear and the conclusion is thought-provoking, one could say shocking: wealth and job creation have diverged. Furthermore, I expect this divergence to grow.

Let’s look at manufacturing, an area with which I have considerable familiarity. – I started my involvement with manufacturing at age 14 as an unskilled laborer in a machine factory in the early 1950s, worked as a technician for a small privately-owned home heaters manufacturer in the late 1950s; later I was Director of Manufacturing Technology Research at a major domestic electronics manufacturer. Following that, I worked for a major international consumer electronics manufacturer where I had full technical responsibility for products and manufacturing processes of a product family producing over a billion-dollar yearly sales. I have intimate familiarity over more than half a century with factory operations, their evolution, and technology. I have built experimental production lines as well as small volume and large volume mass production plants.

Without exception, in high-volume manufacturing, the one never-ending objective is to reduce costs. All successful products deliver a set of consumer benefits at a cost below that available to the consumer from the competition. This is a fundamental truth under all circumstances: a monopoly or a communist state eliminates the competition; in a capitalistic competitive economy, the manufacturers’ emphasis must be to minimize the cost of producing the set of benefits offered to the customer. There are three fundamental cost components: capital, materials, and labor. Once a design is fixed, the materials content is defined and neither capital nor labor can be traded for reduced materials content without changing the design. (Materials costs of course are directly affected by outsourcing versus vertical integration business strategies.) Significant opportunities exist in most manufacturing operations to substitute labor and capital for each other. During experimental and pilot phases capital commitments are likely to be minimized and production is more labor-intensive than during the volume production mature phase. In general, given the state of technology, if automating an operation is feasible, the long-term cost of automation is less than the cost of the substituted labor would be. In advanced economies, if it is technically feasible, the per-unit depreciation cost of the automated equipment, i.e. per unit cost of the capital, tends to be lower than the per-unit labor cost replaced by automation. (In developing economies, for example in China in the recent past, the cost of available capital versus the availability of plenty of low-cost labor may temporarily lead to a lower level of automation than in a contemporary advanced economy; nevertheless, the long term historic trend towards automation is everywhere similar.)

Throughout the history of manufacturing, from the beginning of the industrial revolution through the invention of assembly lines, and into the era of modern chip manufacturing that enables the information age, the relentless drive has been to reduce per unit labor costs. This is true whether the unit is a car (sold by the millions for tens of thousands of dollars each) or an electronic switch (of which there may be billions on chips sold by the billions each for a few dollars or less). Since a unit switch costs less than about a billionth of a dollar, the range of the values of the expected and continuous unit labor cost reductions is astronomical. It has been successfully addressed nonstop for over two hundred years by some of the best brains using every conceivable tool in all known manufacturing industries.

Much of manufacturing technology involves shaping and joining parts from incoming materials, i.e. parts manufacturing and assembly. The incoming materials themselves are likely to have been prepared by shaping and joining at a supplier’s manufacturing plant, and so forth. This supply chain goes back all the way to the extraction and preparation of raw materials that are mostly chemical in nature. Chemical plants have been highly automated already long before the technologies became available for significant automation of the shaping and joining activities. Automation of parts manufacturing also predates large-scale automation of assembly operations.

Automation of assembly operations in its earlier phase employed “hard automation”, where unique machinery needed to be designed for the joining of subassembly elements. This was expensive and slow to develop and at times required a redesign of the incoming elements to facilitate their automated handling. More recently with the availability of robots with more and more articulated motion and grasping capabilities, assembly line designs incorporated more and more robots. In general robotic assembly lines are more flexible than those using hard automation, they can easily be programmed to accommodate the handling of a variety of parts and finished assemblies. By around the year 2000 in mass production lines, human labor was retained principally only in those operations that required eye-hand cooperation. By the second decade of the 21st-century machine vision has become sufficiently well developed to eliminate much of the earlier hand-eye coordination requirement constraint on automation. In fact, machine vision-aided intelligent robots mostly surpass human capability in joining parts: they see better, do not tire, and are more precise. An interesting comparison of human and robotic manufacturing was made in a front-page article in the New York Times on August 19, 2012. According to that article, at a factory in China, hundreds of manual laborers assemble high-end electric shavers; a sister factory in the Netherlands uses 128 robots to complete the same task, without coffee breaks working 24/7 365 days a year. If necessary, in addition to vision, robots in various applications may utilize other available sensors and measuring devices (e.g. tactile, weight, distance) thus making those easily programmable human-like task-oriented performers. There is a considerable public discussion regarding the repatriation of manufacturing that went off-shore; while this is a worthwhile objective to strengthen the national economy, such repatriation is not likely to have a significant impact on recreating the jobs that went overseas. In fact, those jobs are gone forever and the repatriation of some (or even much) manufacturing may occur as part of the natural economic evolution: in the world of robotic manufacturing the cost advantage offered by low-cost overseas labor is gone and the lowest cost may be achievable by locating factories closer to the markets.

The labor content reduction is not confined to manufacturing. In agriculture, we have gone in the course of about 250 years from virtually everyone working the land to a small fraction now so employed. Much of this reduction is the result of large-scale farming with large cultivating machines. In the next phase, farming machinery is moving towards GPS-controlled autonomous equipment. A large scale dairy farm (thousands of heads herd) I visited in Arizona has used milking machines for many years: as the cows arrived at the machine their tags were read by scanners, humans connected and disconnected the machine to the udders, and the volume and quality of the milk were computer analyzed and stored. More recently, in New York family farms (hundred heads herd) started using fully robotic milking and feeding (no farmhands involved at all), obtaining apparently superior results by being able to allow the cows to elect how many times and when they are to be milked by the robots. (N Y Times, Apr. 23, 2014)

According to the U. S. Bureau of Labor Statistics the combined goods-producing industries (mining plus construction plus manufacturing plus agriculture) in 2000, 2010, and 2020 respectively employed, or are forecast to employ, 18.4, 13.9, and 13.1 percent of the working population of the USA.

Activities to save labor through automation are not confined to the goods-producing industries. Such activities can be found in the service industries too and virtually no industry sector is immune to their expected impact. Surgeons now can avail themselves of robotic tools that allow remote operations that save scrubbing and dressing time. No doubt high precision invasive procedures will be more accurately (and thus more safely) performed by robotic devices than by knives held by human hands. Self-driving cars and robots that climb ladders are already operational. Automated warehouse operations are well established. Thus the delivery of online ordered goods from automated warehouse to home via fully autonomous means can be expected soon.

Robotic delivery in hospitals and in homes of personal services for the sick and the aged are being considered. More generally I can envision a world where most of us will have one or more personal assistant robots. Recently a friend visiting Carnegie Mellon University was guided by a Co-Bot; this device is a post on wheels with a screen on top to communicate; it knows its way around, and asks people to push the elevator buttons to get where it needs to go; in addition to guiding people around it picks up the mail and delivers small packages as requested; if stuck somewhere without help, it contacts its headquarters. Walking, seeing, listening, tactile-and force-sensing robots with two legs and two arms already have been demonstrated, they can be programmed to execute various tasks by initially guiding their movements through task-required paths and their actions directed by voice command using well-established speech recognition tools. The natural extension of these processes takes us out of science fiction into the new real-world of the not too distant future.

Already today certain easily programmed two-armed stationary robots can be purchased for $25,000. The inherent cost of an autonomous walking robot is not higher than that of an automobile, in fact, it is probably much lower. Weight is an important cost scaler and a personal assistant robot would certainly weigh much less than a low-cost car. Thus, once the technology is fully developed and demand established, highly automated mass production techniques are applied to the fabrication of robots themselves, a mass market will open up where personal assistant robots will be desired and affordable by all.

Thus robots eliminate jobs. They also may enhance the average person’s quality of life. But with no job, how will a person who earned a modest income and accumulated no wealth pay for his/her necessities that possibly (even may need to) include an intelligent machine personal assistant?


Where we are coming from and are going to

As by now it is well known, and also discussed in some quantitative detail in the Appendix, labor’s share of the slowly growing economy is continuously decreasing. While some economists may argue whether capital and labor are complementary or substitutional, in the practice of manufacturing automation definitely is substituted for labor. Automation is a capital investment, thus capital is substituted for labor. One may argue that with automation the unit costs are reduced, thus prices can be reduced and maybe more units can be sold. This then can lead to the opening of more manufacturing lines to satisfy the growing demand and thus labor is not necessarily displaced; however, when supply already exceeds demand, automation directly displaces labor.

During the 19th and much of the 20th centuries, as one industry matured another new industry was born. For example, the textiles, automobile, and consumer electronics industries evolved consecutively, each creating reasonably well-paying new employment opportunities. In the last quarter of the 20th century, two new industries emerged: microelectronics and information technology (IT). Neither has been a major job creator for the average person. Microelectronics by its very nature needed to be highly automated from the beginning and ultimately displaced earlier consumer electronics products. For example, cheaper and less labor-intensive heavily microelectronics-based HDTV Flat Panels completely replaced the earlier CRT TVs. Both microelectronics and IT provided many new jobs to a new generation of highly trained professionals, but in general, only reduced the employment opportunities of persons without a college education. (See the Appendix for data on the correlation of education with unemployment and income.) 

We face a major crisis. We need to rethink some of our basic economic beliefs; otherwise our relative social peace and our capitalistic society are likely to break down.

Till the 1970s we had a social contract that maintained a balance between capital and labor, whereby capital and labor shared the benefits of productivity gains. Later on, increasing automation, deregulation, and the replacement of the old rule of “one citizen one vote” with the new rule of “one dollar one vote”, ended the social contract. We now have “inequality” that has been growing continuously for well over a quarter of a century and it appears to continue to grow unstoppably in the foreseeable future. Who are the expected winners in our brave new world? They belong to three groups: owners of capital, the top managers of major industrial and financial organizations, and innovators. Some innovators who recently became super-rich have come up with new ideas that appeal to many people and out of this appeal created billion-dollar companies that employ virtually no one.

We have growing inequality and at the same time our machines can produce enough to satisfy all of our physical needs, provided we can pay. Of course, if we cannot pay, because there are no jobs, the machines too will stop and no one’s needs can be satisfied. The key point is that we have more capacity than demand now. In the future, we are likely to continue to employ increasingly robotics and IT technologies in virtually all industries and not only in manufacturing. Then, if we continue on the present course, we will have fewer jobs, and fewer people able to pay for what they need (and what we can easily produce/supply to fulfill their needs). This situation will lead to human suffering, ever-increasing inequality, political unrest, and economic disaster.

What needs to be done?  First, we need to recognize and accept that a significant fraction of society is unemployable. Simply put, we have more people than are needed to produce all the goods and services consumed by our society. We must take care of all. We also need to recognize that no profession is immune to the effects of advances in robotics and IT. Many aspects of law, medicine, education, engineering, human services, etc… are all likely to require reduced human contribution for the completion of given tasks. In education, for example, ongoing experiments in computer-based language teaching methods may initially only match, but then are expected to surpass the effectiveness of classroom teachers.

Once we recognize the above reality, the next step is a new social contract whereby all will share in the productivity gains brought by intelligent machines. We need to reduce the number of working hours; a reduction long overdue. During the Dickensian period of industrial evolution workweeks of 60 hours or even more were not uncommon. By the late 19th century the 48-hour workweek was established. In 1926 Ford introduced the 5-days 40-hours work week. (It is interesting to note that in the Soviet satellite Communist countries the 6-days 48-hours workweek was continued till the late 1960s.) The 5-days 40-hours work week (with 0.5 hours paid lunch break typically bringing the weekly hours worked to 37.5 hours), with minor variations, is still the basic American standard today. Now is the long-overdue time to begin progressive further reductions.

As a social policy, it is better to employ more and have fewer people on welfare. Without a workweek reduction, we will have more people on welfare and fewer employed. As a matter of politics, to accomplish this in the current environment is very difficult. But sooner or later it must be done, and the sooner it is done, the better it is for all. Clearly, to support all of us working fewer hours, will require some form of taxation and transfer payments. Clearly, we need a redistribution of income and possibly wealth also. Clearly, some may think of this as socialistic. So be it! To save our free enterprise capitalistic system, we must update it; better to adjust and fine-tune it than to lose it altogether.


Concluding remarks: a Global perspective

Thus far I have only discussed the specific issues from a US perspective; inequality and unemployment are global issues. In the developing world “rent-seeking” crony capitalism rules, where political connections created a super-rich class, typically closely entwined with the political elite. Russia and Ukraine are prime examples of this situation, as well as is China. Inequality in these countries far exceeds that in the USA.  In general, in the course of the last decades, inequality has grown around the world. Nevertheless, now inequality in the USA exceeds that in all other countries of the developed world. According to the CIA World Factbook, the least inequality, as measured by the Gini index, exists in Sweden with rank order 139 out of 139 countries examined while the USA is assigned position 41 in the Global rank order. USA inequality is worse than that in many African countries; it is also worse than in Russia and in Ukraine, but better than in China.  In several countries of the developed world, some progress has been made in addressing the fundamental unemployment issue. In France, the workweek has been reduced to 35 hours. In Germany to ease unemployment “Kurzarbeit” (short-work) is widely practiced, whereby instead of lay-offs reduced working hours have been deployed with government subsidies providing partial compensation for the lost wages. A typical 40 hours/week American worker with a 0.5-hour lunch break, 10 days of paid vacation, and 10 paid holidays would work a total of 1,800 hours per year; on a 52-week basis thus Americans work on the average about 34.5 hours per week. As a result of “Kurzarbeit” and also of more generous vacation plans, in 2011 Germans worked 1,330 hours in the year, or on the average 25.6 hours per week. In 2011 the US unemployment rate was about 9% and the German was about 6%. In 2014 at the time of this writing, the US unemployment rate is about 6.7%, and that in Germany was 5.1%. Several countries in the EU are ahead of the USA in addressing both inequality and unemployment.




Appendix: Quantification of American income and wealth trends.

The average person’s income and wealth are represented by the median income and wealth. The relative value of the median versus the mean provides information about the distribution across the population. While banks and the governments compile huge amounts of data, in no single source could I find data for the four numbers: mean and median income and wealth on a given recent date (or tax year). From The US Census Bureau, “Income, Poverty, and Health Insurance Coverage in the United States: 2012”, p.6, Table 1 for 2011 2011 and 2012 gives the median household income as $51k (number of households in 2011 121 million and in 2012 122 million). From the US Census Bureau, “Historical Income Tables: Households, Table H-3. Mean Household Income Received by Each Fifth and Top 5 Percent” I compute the mean household incomes for 2012 as $71k and for 2011 as $70k. From The US Census Bureau, “Household Wealth in the U.S.: 2000 to 2011”, the median net worth of US households was $69k; the same source gives the aggregate net worth of all US households in 2011 as $40.2 Trillion from which for 121 million households I compute a mean household net worth as $330k. [There are surprisingly large differences in the data available from different sources. For example, from a Congressional Research Service publication (fas.org/sgp/crs/misc/RL33433.pdf) I found that based on data from a Federal Reserve Survey of Consumer Finances (SCF), mean household net worth was $498,800 and median household net worth was $77,300 in 2010.” In any case, the general message is clear; the mean net worth is much larger than the median, indicating that an affluent minority owns most of the wealth in the USA.]

Thus the Census data for 2011 shows that the median family income and net worth are substantially lower than the corresponding means ($51k vs. $70k and $69k vs. $330k, respectively). Furthermore, the data also shows that between 2011 and 2012 the median family stayed the same while the mean increased. Through the years examined, we can see the much-discussed fact that the average American household is not sharing the benefits of the growing US economy.

From The US Census Bureau, “Historical Income Tables: Households, Table H-6. Regions-by Median and Mean Income”, in the course of 25 years from 1986 and 2011 the inflation-adjusted median household income increased by 2.5%, while the corresponding mean by 15.9%. During the same period, the inflation-adjusted GDP per capita increased by close to 50% (45.2% according to indexmundi.com/facts/united-states/gdp-per-capita48.1% according to multpl.com/us-real-gdp-per-capita/table/by-year, and 47.9% according to ers.usda.gov/datafiles/International_Macroeconomic_Data/Historical_Data_Files/HistoricalRealGDPValues.xls). During the same period total government expenditures as a percent of the GDP changed little (was about 35% according to usgovernmentspending.com). Also in 2011 the GDP per capita was $50k and I compute $129k GDP per household (311.6 million people, 121 million households, 2.58 people per household); if I adjust this to split 65/35 between households and the government, I calculate $84k as the mean value of the households’ share of the GDP. Since the mean household income was $70k (see above), I conclude that about 17% is accrued somewhere, not contributing to family incomes. Also, we conclude that the GDP increase is not reflected in the increase in incomes, or in increased government expenditures. Where does it go? -- Both in the short term and in the long term the income of the affluent increased faster than that of the household of the average American.

The evolution of Family Net Worth distribution in the USA over a quarter of a century is shown in the table below (reproduced from 2.ucsc.edu/whorulesamerica/power/wealth.html) for three income percentile groups.
           Top 1%    Next 19%    Bottom 80%
1983       33.8%      47.5%        18.7%
1989       37.4%      46.2%        16.5%
1992       37.2%      46.6%        16.2%
1995       38.5%      45.4%        16.1%
1998       38.1%      45.3%        16.6%
2001       33.4%      51.0%        15.6%
2004       34.3%      50.3%        15.3%
2007       34.6%      50.5%        15.0%
2010       35.4%      53.5%        11.1%

We conclude that during the 27-year period shown above, the lower 80% saw close to a factor two reduction in its share of the Nation’s wealth. Furthermore, according to the Congressional Research Service (CRS) (see fas.org/sgp/crs/misc/RL33433.pdf, Table 2), from 1989 through 2010, the upper 10% of wealth owners (here not income percentile but wealth percentile) increased its share of the total net worth from 67.2 % to 74.5 %.

From the above referenced CRS report Table 3, reproduced here below, additional income and net worth data are available. It is interesting to note that up to the 90th percentile, in the ranges shown the income distribution is flat, medians and means are essentially the same as one would expect for the narrow percentile ranges selected. However, in the range of 90% to 100%, the mean is substantially larger than the median. Here the small elite group of extremely high earners skews even the narrow high-income percentile group’s income average strongly upwards. It is also noteworthy that the lower the income group, the higher is the mean to the median ratio for net worth; this suggests that in the lower-income brackets some households manage to accrue significantly more wealth, obviously not by earning more, but presumably by saving and/or inheriting more.

Household Income and Net Worth by Income Class (2010 dollars) – Source: CRS Report for Congress www.crs.gov RL33433










Income
      Income ($k)
   Net worth ($k)
% Households
Percentile
Median
Mean
Median
Mean
Who Saved
All
45.80
78.50
77.38
498.80
52.00
<  20%
13.40
12.90
6.20
116.80
32.30
20% to 40%
28.10
27.90
25.60
127.90
43.40
40% to 60%
45.80
46.30
65.90
199.00
49.80
60% to 80%
71.70
73.60
128.60
293.90
60.10
80% to 90%
112.80
114.60
286.60
567.20
67.70
90% to 100% 
205.30
349.00
1,194.30
2,944.10
80.90
Note: Income data are for 2009, the year before the 2010 SCF was conducted survey. Asset and liability data are for 2010, as of the time interviews were conducted.

The foregoing quantitative details clearly indicate that in our era the rich get richer and a few may get rich, but the majority is left behind economically and likely heading towards marginalization. – Nevertheless, education pays. The table below is reproduced from the website of the U.S. Department of Labor, U.S. Bureau of Labor Statistics.
Education attained
Unemployment rate in 2013 (Percent)
Median weekly earnings ($)
Doctoral degree
2.2
1,623
Professional degree
2.3
1,714
Master's degree
3.4
1,329
Bachelor's degree
4.0
1,108
Associate's degree
5.4
777
Some college, no degree
7.0
727
High school diploma
7.5
651
Less than a high school diploma
11.0
472
Note: Data are for persons age 25 and over. Earnings are for full-time wage and salary workers.
Source: Current Population Survey, U.S. Department of Labor, U.S. Bureau of Labor Statistics

In fact, the benefit of education in terms of a low unemployment rate is even greater than the above table suggests. In lower educated segments of the population, the participation rate (fraction seeking employment) is lower than in the better-educated segments. Presumably, because a higher fraction of less-educated people has given up looking for jobs than did the more-educated, rather than low earners not needing the income. The table below also reproduced here from the website of the U.S. Department of Labor, U.S. Bureau of Labor Statistics (it again covers persons 25 years and over) shows that in the lowest educational category fewer than half participate (want jobs) while at the high education levels more than three-quarters do so. It is interesting and for me surprising to note that in all education categories the participation rate decreased from the year 2012 to 2013. The reason for this is not clear to me. Could it be a statistical effect of an aging population?

According to the Census Bureau (Educational Attainment in the United States: 2013 - Detailed Tables) out of a total population over 25 years of age, 12% hold graduate and/or professional degrees. Many, if not most of these advanced degree holders may not be super-rich but are in the upper 20% of US earners and wealth holders.