Tag Archives: Scalability

Inequality and Inequality Propaganda

21 Saturday Dec 2019

Posted by in Income Distribution, Inequality, Uncategorized

≈ Leave a comment

Tags

, , , , , , , , , , , , , , , , , , , , , , , , , ,

I’m an “inequality skeptic”, first, with respect to its measurement and trends; and second, with respect to its consequences. Economic inequality in the U.S. has not increased over the past 60 years as often claimed. And some degree of ex post inequality, in and of itself, has no implication for real economic well-being at any point on the socioeconomic spectrum, the growls of class-warmongers aside. So I’m not just a skeptic. I’m telling you the inequality narrative is BS! The media has been far too eager to promote distorted metrics that suggest widening disparities and presumed injustice. Left-wing politicians such as Bernie Sanders, Elizabeth Warren, and Alexandra Ocasio-Cortez pounce on these reports with opportunistic zeal, fueling the flames of class warfare among their sycophants.

Measurement

Comparisons of income groups and their gains over time have been plagued by a number of shortcomings. Jeff Jacoby reviews issues underlying the myth of a widening income gap. Today, the top 1% earns about the same share of income as in the early 1960s, according to a recent study by two government economists, Gerald Auten and David Splinter.

Jacoby recounts distortions in the standard measures of income inequality:

  • The comparisons do not account for tax burdens and redistributive government transfer payments, which level incomes considerably. As for tax burdens, the top 1% paid more taxes in 2018 than the bottom 90% combined.
  • The focus of inequality metrics is typically on households, the number of which has expanded drastically with declines in marriage rates, especially at lower income levels. Incomes, however, are more equal on a per capital basis.
  • The use of pension and retirement funds like IRAs and 401(k) plans has increased substantially over the years. The share of stock market value owned by retirement funds increased from just 4% in 1960 to more than 50% now. As Jacoby says, this has “democratized” gains in asset prices.
  • A change in the tax law in 1986 led to reporting of more small business income on individual returns, which exaggerated the growth of incomes at the high-end. That income had already been there.
  • People earn less when they are young and more as they reach later stages of their careers. That means they move up through the income distribution over time, yet the usual statistics seem to suggest that the income groups are static. Jacoby says:

“Contrary to progressive belief, America is not divided into rigid economic strata. The incomes of the wealthy often decline, while many taxpayers go from being poor at one point to not-poor at another. Research shows that more than one-tenth of Americans will make it all the way to the top 1 percent for at least one year during their working lives.”

Mark Perry recently discussed America’s record middle-class earnings, emphasizing some of the same subtletles listed above. A middle income class ($35k-$100k in constant dollars) has indeed shrunk over the past 50 years, but most of that decrease was replaced by growth in the high income strata (>$100k), and the lower income class (<$35k) shrank almost as much as the middle group in percentage terms.

Causes

What drives the inequality we actually observe, after eliminating the distortions mentioned above? The reflexive answer from the Left is capitalism, but capitalism fosters great social and economic mobility relative to authoritarian or socialist regimes. That a few get fabulously rich under capitalism is often a positive attribute. A friend of mine contends that most of the great fortunes made in recent history involve jobs for which the product or service produced is highly scalable. So, for example, on-line software and networks “scale” and have produced tremendous fortunes. Another way of saying this is that the marginal cost of serving additional customers is near zero. However, those fortunes are earned because consumers extract great value from these products or services: they benefit to an extent exceeding price. So while the modern software tycoon is enriched in a way that produces inequality in measured income, his customers are enriched in ways that aren’t reflected in inequality statistics.

Mutually beneficial trade creates income for parties on only one side of a given transaction, but a surplus is harvested on both sides. For example, an estimate of the consumer surplus earned in transactions with the Uber ride-sharing service in 2015 was $1.60 for every dollar of revenue earned by Uber! That came to a total of $18 billion of consumer surplus in 2015 from Uber alone. These benefits of free exchange are difficult to measure, and are understandably ignored by official statistics. They are real nevertheless, another reason to take those statistics, and inequality metrics, with a grain of salt.

Certain less lucrative jobs can also scale. For example, the work of a systems security manager at a bank produces benefits for all customers of the bank, and at very low marginal cost for new customers. Conversely, jobs that don’t scale can produce great wealth, such as the work of a highly-skilled surgeon. While technology might make him even more productive over time, the scalability of his efforts are clearly subject to limits. Yet the demand for his services and the limited supply of surgical skills leads to high income. Here again, both parties at the operating table make gains (if all goes well), but only one party earns income from the transaction. These examples demonstrate that standard metrics of economic inequality have severe shortcomings if the real objective is to measure differences in well-being. 

Economist Robert Samuelson asserts that “workaholics drive inequality“, citing a recent study by Edward E. Leamer and J. Rodrigo Fuentes that appeals to statistics on incomes and hours worked. They find the largest income gains have accrued to earners with high educational attainment. It stands to reason that higher degrees, and the longer hours worked by those who possess them, have generated relatively large income gains. Samuelson also cites the ability of these workers to harness technology. So far, so good: smart, hard-working students turn into smart, hard workers, and they produce a disproportionate share of value in the marketplace. That seems right and just. And consumers are enriched by those efforts. But Samuelson dwells on the negative. He subscribes to the Ludditical view that the gains from technology will accrue to the few:

“The Leamer-Fuentes study adds to our understanding by illuminating how these trends are already changing the way labor markets function. … The present trends, if continued, do not bode well for the future. If the labor force splits between well-paid workaholics and everyone else, there is bound to be a backlash — there already is — among people who feel they’re working hard but can’t find the results in their paychecks.“

That conclusion is insane in view of the income trends reviewed above, and as a matter of economic logic: large income gains might accrue to the technological avant guarde, but those individuals buy things, generating additional demand and income gains for other workers. And new technology diffuses over time, allowing broader swaths of the populace to capture value both in consumption and production. Does technology displace some workers? Of course, but it also creates new, previously unimagined opportunities. The history of technological progress gives lie to Samuelson’s perspective, but there will always be pundits to say “this time it’s different”, and it probably sounds heroic to their ears.

Consequences

The usual discussions of economic inequality in media and politics revolve around an egalitarian ideal, that somehow we should all be equal in an absolute and ex post sense. That view is ignorant and dangerous. People are not equal in terms of talent and their willingness to expend effort. In a free society, the most talented and motivated individuals will produce and capture more value. Attempts to make it otherwise can only interfere with freedoms and undermine social welfare across the spectrum. This post on the Declination blog, “The Myth of Equality“, is broader in its scope but makes the point definitively. It quotes the Declaration of Independence:

“We hold these truths to be self-evident, that all men are created equal, that they are endowed by their Creator with certain unalienable Rights, that among these are Life, Liberty and the Pursuit of Happiness.”

The poster, “Thales”, goes on to say:

“The context of this was within an implied legal framework of basic rights. All men have equal rights granted by God, and a government is unjust if it seeks to deprive a man of these God-given rights. … This level of equality is both the basis for a legal framework limiting the power of government, and a reference to the fact that we all have souls; that God may judge them. God, being omniscient, can be an absolute neutral arbiter of justice, having all the facts, and thus may treat us with absolute equality. No man could ever do this, though justice is often better served by man at least making a passing attempt at neutrality….”

Attempts to go beyond this concept of ex ante equality are doomed to failure. To accept that inequalities must always exist is to acknowledge reality, and it serves to protect rights and opportunities broadly. To do otherwise requires coercion, which is violent by definition. In any case, inequality is not as extreme as standard metrics would have us believe, and it has not grown more extreme.

Central Planning Fails to Scale, Unlike Spontaneous Order

05 Tuesday Jun 2018

Posted by in Central Planning, Markets, Price Controls

≈ Leave a comment

Tags

, , , , , , , , , , , , , ,

The proposition that mankind is capable of creating a successful “planned” society is at least as old as the Bronze Age. Of course it’s been tried. The effort necessarily involves a realignment of the economic and political landscape and always requires a high degree of coercion. But putting that aside, such planning can never be successful relative to spontaneous order of the kind that dominates private affairs in a free society. The task of advancing human well-being given available resources has never been achieved under central planning. It always fails miserably in this regard, and it always will fail to match the success of decentralized decision-making and private markets.

There are various ways to explain this fact, but I recently came across an interesting take on the subject having to do with the notion of scalability. Francis Turner offers this note on the topic at the Liberty.me blog. To begin, he gives a lengthy quote from a software developer who relates the problems of social and economic planning to the complexity of managing a network. On the topic of scale, the developer notes that the number of relationships in a network increases with the square of the number of its “nodes”, or members:

“2 nodes have 1 potential relationship. 4 nodes (twice as many) has 6 potential relationships (6 times as many). 8 nodes (twice again) has 28 potential relationships. 100 nodes => [4,950] relationships; 1,000 nodes => 499,500 relationships—nearly half a million.“

Actually, the formula for the number of potential relationships or connections in a network is n*(n-1)/2, where n is the number of network nodes. The developer Turner  quotes discusses this in the context of two competing network management structures: client-server and peer-to-peer. Under the former, the network is managed centrally by a server, which communicates with all nodes, makes various decisions, and routes communications traffic between nodes. In a peer-to-peer network, the work of network management is distributed — each computer manages its own relationships. The developer says, at first, “the idea of hooking together thousands of computers was science fiction.” But as larger networks were built-out in the 1990s, the client-server framework was more or less rejected by the industry because it required such massive resources to manage large networks. In fact, as new nodes are added to a peer-to-peer network, its capacity to manage itself actually increases! In other words, client-server networks are not as scalable as peer-to-peer networks:

“Even if it were perfectly designed and never broke down, there was some number of nodes that would crash the server. It was mathematically unavoidable. You HAVE TO distribute the management as close as possible to the nodes, or the system fails.

… in an instant, I realized that the same is true of governments. … And suddenly my coworker’s small government rantings weren’t crazy…”

This developer’s epiphany captures a few truths about the relative efficacy of decentralized decision-making. It’s not just for computer networks! But in fact, when it comes to network management, the task is comparatively simple: meet the computing and communication needs of users. A central server faces dynamic capacity demands and the need to route changing flows of traffic between nodes. Software requirements change as well, which may necessitate discrete alterations in capacity and rules from time-to-time.

But consider the management of a network of individual economic units. Let’s start with individuals who produce something… like widgets. There are likely to be real economies achieved when a few individual widgeteers band together to produce as a team. Some specialization into different functions can take place, like purchasing materials, fabrication, and distribution. Perhaps administrative tasks can be centralized for greater efficiency. Economies of scale may dictate an even larger organization, and at some point the firm might find additional economies in producing widget-complementary products and services. But eventually, if the decision-making is centralized and hierarchical, the sheer weight of organizational complexity will begin to take a toll, driving up costs and/or diminishing the firm’s ability to deal with changes in technology or the market environment. In other words, centralized control becomes difficult to scale in an efficient way, and there may be some “optimal” size for a firm beyond which it struggles.

Now consider individual consumers, each of whom faces an income constraint and has a set of tastes spanning innumerable goods. These tastes vary across time scales like hour-of-day, day-of-week, seasons, life-stage, and technology cycles. The volume of information is even more daunting when you consider that preferences vary across possible price vectors and potential income levels as well.

Can the interactions between all of these consumer and producer “nodes” be coordinated by a central economic authority so as to optimize their well-being dynamically, subject to resource constraints? As we’ve seen, the job requires massive amounts of information and a crushing number of continually evolving decisions. It is really impossible for any central authority or computer to “know” all of the information needed. Secondly, to the software developer’s point, the number of potential relationships increases with the square of the number of consumers and producers, as does the required volume of information and number of decisions. The scalability problem should be obvious.

This kind of planning is a task with which no central authority can keep up. Will the central authority always get milk, eggs and produce to the store when people need it, at a price they are willing to pay, and with minimal spoilage? Will fuel be available such that a light always turns on whenever they flip the switch? Will adequate supplies of medicines always be available for the sick? Will the central authority be able to guarantee a range of good-quality clothing from which to choose?

There has never been a central authority that successfully performed the job just described. Yet that job gets done every day in free, capitalistic societies, and we tend to take it for granted. The massive process of information transmission and coordination takes place spontaneously with spectacularly good results via private discovery and decision-making, secure property rights, markets, and a functioning price mechanism. Individual economic units are endowed with decision-making power and the authority to manage their own relationships. And the spontaneous order that takes shape remains effective even as networks of economic units expand. In other words, markets are highly scalable at solving the eternal problem of allocating scarce resources.

But thus far I’ve set up something of a straw man by presuming that the central authority must monitor all individual economic units to know and translate their demands and supplies of goods into the ongoing, myriad decisions about production, distribution and consumption. Suppose the central authority takes a less ambitious approach. For example, it might attempt to enforce a set of prices that its experts believe to be fair to both consumers and producers. This is a much simpler task of central management. What could go wrong?

These prices will be wrong immediately, to one degree or another, without tailoring them to detailed knowledge of the individual tastes, preferences, talents, productivities, price sensitivities, and resource endowments of individual economic units. It would be sheer luck to hit on the correct prices at the start, but even then they would not be correct for long. Conditions change continuously, and the new information is simply not available to the central authority. Various shortages and surpluses will appear without the corrective mechanism usually provided by markets. Queues will form here and inventories will accumulate there without any self-correcting mechanism. Consumers will be angry, producers will quit, goods will rot, and stocks of physical capital will sit idle and go to waste.

Other forms of planning attempt to set quantities of goods produced and are subject to errors similar to those arising from price controls. Even worse is an attempt to plan both price and quantity. Perhaps more subtle is the case of industrial policy, in which planners attempt to encourage the development of certain industries and discourage activity in those deemed “undesirable”. While often borne out of good intentions, these planners do not know enough about the future of technology, resource supplies, and consumer preferences to arrogate these kinds of decisions to themselves. They will invariably commit resources to inferior technologies, misjudge future conditions, and abridge the freedoms of those whose work or consumption is out-of-favor and those who are taxed to pay for the artificial incentives. To the extent that industrial policies become more pervasive, scalability will become an obstacle to the planners because they simply lack the information required to perform their jobs of steering investment wisely.

Here is Turner’s verdict on central planning:

“No central planner, or even a board of them, can accurately set prices across any nation larger than, maybe, Liechtenstein and quite likely even at the level of Liechtenstein it won’t work well. After all how can a central planner tell that Farmer X’s vegetables taste better and are less rotten than Farmer Y’s and that people therefore are prepared to pay more for a tomato from Farmer X than they are one from Farmer Y.”

I will go further than Turner: planning can only work well in small settings and only when the affected units do the planning. For example, the determination of contract terms between two parties requires planning, as does the coordination of activities within a firm. But then these plans are not really “central” and the planners are not “public”. These activities are actually parts of a larger market process. Otherwise, the paradigm of central planning is not merely unscalable, it is unworkable without negative consequences.

Finally, the notion of scalability applies broadly to governance, not merely economic planning. The following quote from Turner, for example, is a ringing endorsement for federalism:

“It is worth noting that almost all successful nations have different levels of government. You have the local town council, the state/province/county government, possibly a regional government and then finally the national one. Moreover richer countries tend to do better when they push more down to the lower levels. This is a classic way to solve a scalability problem – instead of having a single central power you devolve powers and responsibilities with some framework such that they follow the general desires of the higher levels of government but have freedom to implement their own solutions and adapt policies to local conditions.”