The myth of dynastic wealth: the rich get poorer.

• Author: Arnott, Robert

Thomas Piketty's Capital in the Twenty-First Century rocketed to the top of the best-seller lists the moment it was published in 2013, and remained there for months. While this feat is quite remarkable for a weighty tome on economics, it's no mystery why Piketty's magnum opus created such a sensation; it is clearly articulated, is accessible to the non-economist, and contains a trove of historical insights.

We believe Piketty's core message is provably flawed on several levels, as a result of fundamental and avoidable errors in his basic assumptions. (1) He begins with the sensible presumption that the return on invested capital, r, exceeds macroeconomic growth, g, as must be true in any healthy economy. But from this near-tautology, he moves on to presume that wealthy families will grow ever richer over future generations, leading to a society dominated by unearned, hereditary wealth. Alas, this logic holds true only if the wealthy never dissipate their wealth through spending, charitable giving, taxation, ill-advised investments, and splitting bequests among multiple heirs. As individuals, and as families, the rich generally do not get richer: after a fortune is first built, the rich often get relentlessly and inexorably poorer.

The evidence Piketty uses in support of his thesis is largely anecdotal, drawn from the novels of Austen and Balzac, and from the current fortunes of Bill Gates and Liliane Bettencourt. If Piketty is right, where are the current hyper-wealthy descendants of past entrepreneurial dynasties--the Astors, Vanderbilts, Camegies, Rockefellers, Mellons, and Gettys? Almost to a man (or woman) they are absent from the realms of the super-affluent. Our evidence--used to refute Piketty's argument--is empirical, drawn from the rapid rotation of the hyper-wealthy through the ranks of the Forbes 400, and suggests that, at any given time, half or more of the collective worth of the hyper-wealthy is first-generation earned wealth, not inherited wealth.

The originators of great wealth are one-in-a-million geniuses; their innovation, invention, and single-minded entrepreneurial focus create myriad jobs and productivity enhancements for society at large. They create wealth for society, from which they draw wealth for themselves. In contrast, the descendants of the hyper-wealthy rarely have that same one-in-a-million genius. Bettencourt, cited by Piketty, is a clear exception. Typically, we find that descendants halve their inherited wealth--relative to the growth of per capita GDP--every 20 years or less, without any additional assistance from Piketty's redistribution prescription.

Dynastic wealth accumulation is simply a myth. The reality is that each generation spawns its own entrepreneurs who create vast pools of entirely new wealth, and enjoy their share of it, displacing many of the preceding generations' entrepreneurial wealth creators. Today, the massive fortunes of the 19th century are largely depleted and almost all of the fortunes generated just a half-century ago are also gone. Do we really want to stifle entrepreneurialism, invention, and innovation in an effort to accelerate the already-rapid process of wealth redistribution?

Piketty's Core Thesis

The central thesis of Piketty's book is his prediction that the problem of unequal wealth distribution will worsen in the 21st century, further exacerbating the economic divisions between society's haves and have-nots. This is where his biases lead him astray. We agree that inequality in wealth distribution has intensified in the recent past, but we dispute Piketty's rationale for its source and its persistence, as well as his contention that it is a problem in need of fixing. Piketty devotes the largest portion of his book to a treasure trove of historical and geographical time-series on income and wealth distributions. Curiously, however, he ignores his own data, basing his predictions and prescriptions primarily on what he calls the two "fundamental laws of capitalism."

The first "law" states that capital's contribution to national income, denoted as [alpha], can be expressed as

(1) [alpha] = r[beta]

where r is the return on capital and [beta] is the capital-to-income ratio, a measure of the capital intensity in a given society. This relationship is intuitively obvious; if a nation's capital is 600 percent of national income (or the equivalent of six years of national income), and if the real return on that capital is 5 percent, then the portion of national income earned from capital will be 30 percent.

The classical estimate of capital's contribution to national income is 10 percent to 15 percent. Piketty's estimate of a is higher, for two reasons. First, his definition of capital is very broad, including not only investment capital, but also productive industrial assets, real estate, and even public assets, such as highways and national parks. Second, and more importantly, Piketty's approximation of r, while perhaps correct in the distant past, is irrelevant in today's low-yield world--but more on that later.

Although Piketty's first fundamental law is self-evident (it is, after all, purely a definition), his second "law" is less intuitive. The ratio of capital to income, [beta], is defined by:

(2) [beta] = s/g

where s is the savings rate and g the real growth rate of the aggregate economy. For example, with a savings rate of 12 percent and an economic growth rate of 2 percent a year, long-run capital accumulation will eventually expand to 600 percent of national income. The higher the savings rate and/or the lower the labor force growth and/or the lower the productivity growth, the higher the [beta]. The equation expresses a dynamic whose consequences can play out over decades.

The relationships represented in Piketty's second law have important implications in a world of slower macroeconomic growth. Our research, and the research of many others, suggests that future economic growth will slow considerably due to two demographic headwinds. First, labor force growth in the developed world has already slowed sharply, even reversing in some developed economies. Second, the global labor force is aging. Because productivity growth occurs at a much faster pace for young adults than for mature adults who are approaching peak levels of productivity, an older workforce means fewer productivity gains. A slowdown in productivity growth is already underway in the developed world, and it is only a matter of time--roughly one generation--before the emerging economies suffer the same fate: sooner in China and Russia, later in India. If growth is halved, then, ceteris paribus, aggregate capital must double as a multiple of income, unless r falls even further than g, either as a consequence of demographic pressures or in parallel with demographic shifts. (2)

Over the last 125 years or so, the global capital-to-income ratio, 3, has ranged between a low of about 200 percent following the devastation wrought by two world wars and a global depression, and today's high of roughly 600 percent, the level also reached during the belle epoque, or robber baron, period of the late 19th century. Piketty asserts p is going even higher from here, to 700 percent by the year 2100. (3) His claim is based on his estimates of the factors of P--economic growth and the savings rate. He forecasts that annual real growth in global output will decline by half from the current

3.0 percent to 1.5 percent in the second half of the 21st century. We agree with his expectations of slower growth. But we have to disagree with his net savings rate assumption of a constant 10 percent. Is his presumption of a fixed savings rate not a rather heroic assumption, given the objective evidence of recent decades?

In Piketty's eyes such a high capital-to-income ratio is cause for alarm. Why so? Piketty is repelled by a high P because a society's total capital, or wealth, is almost always much more unevenly distributed than its income, and a record-high P means a larger pool of wealth will soon be disproportionately enjoyed by the wealthiest members of society. By way of example, Piketty posits that in 2010 the share of earned income of the top 1 percent of American workers was 17.4 percent, whereas the top wealthiest 1 percent owned a

33.8 percent share of total societal wealth.

Elementary finance, though, tells us that this matters little: A doubling of capital wealth due merely to a doubling of valuation, and so to a doubling of the concentration of capital, leaves these 1 percenters with no increase in the income earned on their capital. Incredibly, in Piketty's narrative, the return on capital, r, is insensitive to wealth accumulation, so capital's contribution to income, [alpha], is largely driven by the capital-to-income ratio, (3 (Piketty 2014:206). (4) Indeed, to those of us who manage investments, it evokes incredulity to assert that, if the return on capital were halved, a concurrent doubling of [beta] would not change the respective rewards to labor and capital. In fact, in our view, these precise realignments in rewards can almost entirely explain the growing wealth inequality of the past 30 years.

If slower demographic growth coupled with slower productivity growth would not reduce the return on capital, then Piketty would be correct that an increasing gap between capital returns and growth (r - g) would feed directly into a society's wealth inequality. Because g has been falling in developed nations, Piketty fears that wealth inequality may grow to dangerous levels. His fear--shared by Occupy Wall Street and similar groups--is one manifestation of a perceived widening of the social divide. The increasing attention given to this fear and the socioeconomic agendas fueled by it were important catalysts for our decision to pursue the research we present here, into the objective evidence for or against Piketty's dynastic wealth narrative.

This fear seems to us based on an...