Proposition: Inequality is a drag on global growth

James Radner is Assistant Professor in the School of Public Policy and Governance, University of Toronto, where he also directs the Boreal Institute for Civil Society (for): “That state is opulent where the necessaries and conveniencies of life are easily come at,” wrote Adam Smith in his Lectures on Jurisprudence. In the nearly 250 years since then, humanity has generated an utterly unprecedented explosion of economic product – enough to transform, or even jeopardize, our very biosphere. How have we done against Smith’s criterion for economic success – that is, what he charmingly called “nothing else […] but this comeattibleness?” I wish to argue that inequality has been and remains a terrible drag on growth in comeattibleness.

Start with food and water as two emblematic “necessaries of life.” In this early new century, some 950 million people go to bed hungry each night, and 1.1 billion people have no access to safe drinking water. About half of our little children (under the age of five) are malnourished, such that millions of them die each year. Some 200 million young survivors are likely to suffer lifelong effects from disrupted brain development. Consider necessaries and conveniencies together: 1.2 billion people – a number roughly equal to the entire world population in Smith’s day – live in extreme poverty, with less ‘comeattibleness’ than our hunter-gatherer ancestors enjoyed 10,000 years ago.

This huge, shocking shortfall in comeattibleness is a result of what Lant Pritchett called “divergence, big time” – to wit, the massive inequality of economic development over the two centuries since Smith.  That development, with its enormous increases in production, has yielded far less growth in comeattibleness than it could or should have. But could we not solve this problem, and seemingly eliminate inequality’s drag, simply by producing more for everyone without changing the relative distribution of life’s necessaries and conveniencies?

No. The basic reason is the law of diminishing returns. Those of us now benefiting from huge accumulations of capital and technology can indeed get richer through further investment and innovation, but the rate of increase in production available to us is far less than what is required and feasible in less opulent places. The great escapes from poverty in our time, by hundreds of millions of people – first in Japan, then in the Asian ‘tiger’ economies, then in China – have all involved rapid ‘catch-up,’ as it were, with production rising far faster than in already opulent states. Achieve that where all of today’s poor people live, and world inequality must contract. Thus, whatever mix of production and distribution measures we may favour, we cannot achieve decent growth in comeattibleness without tackling inequality. We need convergence, big time.

Fred Lazar is Associate Professor of Economics in the Schulich School of Business, York University (Toronto) (against): If I knew what comeattibleness really meant and how this relates to economic concepts and values that I understand, I might be able to tackle the arguments head-on. Alas, I do not know what “decent growth in comeattibleness” implies in terms of aggregate economic growth, and indeed why reducing inequality across countries really matters. Why should less inequality enhance comeattibleness? And what really is the root cause of a “shortfall” in comeattibleness?

Setting my ignorance aside, let me start with the observations regarding the extent of absolute poverty today. This poverty is concentrated in countries in Africa and Central Asia, although large numbers can still be found in China. What lies behind the poverty? Answer: Squandered opportunities, resulting from corruption, democratic deficits, and petty tribalism. Will convergence help?

Growth in these countries has, for the most part, been inhibited by internal problems. Will these internal problems be rectified over time? Perhaps, but not in all of these countries, and not within a 20-year timeframe.

Will massive transfers of income and wealth from the ‘have’ countries accelerate the process? Unlikely. The ‘have’ countries do not have any obvious obligations to the ‘have not’ countries. Moreover, international bodies – the UN agencies at the top of the list – have proven themselves incapable of acting as the agents for transfers. They tend to be as corrupt and dysfunctional as the countries that they supposedly try to help. Finally, throwing money at these countries will only increase the payoffs for corruption and kleptocracy, and thus encourage more such behaviour.

One last observation: The law of diminishing returns has been disproved repeatedly since the days of Malthus.

JR: You pack a lot into a small space, with concision that I am afraid I cannot match. For example, I am no expert on interpreting Adam Smith and “comeattibleness.” But I do find an important warning in his insistence that “nothing else can deserve the name of opulence but this comeattibleness […] whether money or other things of that sort abound or not.” Our system of national accounts – summing money to get things like GDP – is a 20th century invention that certainly has instrumental value for macroeconomic management. Indeed, the concept of GDP emerged from the effort to understand the Great Depression. But as imprecise as Smith’s “comeattibleness” may be, it is a far more plausible standard of economic success than GDP.

One way of seeing why this is so is to consider what today’s economists call the marginal utility of wealth. If Bill Gates were gifted $200 a year, it would make no difference in his access to life’s necessaries and conveniencies. If I received such a gift, I could gain some small pleasure – some convenience – from it, but it would not change my life. Yet for the people whom I discussed in my initial comments, such a sum would be transformative. So helpful would it be at ‘coming at’ life’s necessaries that it could easily become the difference between life and death. Thus, while central bankers may appropriately use GDP to help them manage the business cycle, we cannot assess economic gain without considering distribution.

Perhaps an example closer to home would help. From 1980 to 2005, Canada’s real GDP per person grew by nearly 50 percent. In that time, the median real income for the top quintile of full-time Canadian workers grew from C$74,084 to $86,254; the figure for the middle quintile was virtually flat – $41,348 to $41,431 – while the bottom quintile saw a decline from $19,367 to $15,375. To be sure, six numbers (from a small Statistics Canada table) cannot summarize economic performance, but suppose just for illustration that the stylized story that these figures suggest becomes our future (or that of a hypothetical country), without mitigating factors. Is this economic growth? I think not. It amounts to a big increase in human suffering for poor people, and a future of drudgery without progress for middle class families in exchange for a little more comfort for the already comfortable. I interpret that as a chilling loss of comeattibleness in the life of the society as a whole.

Let me also offer some shorter responses on additional points: Our escape from the Malthusian trap is a wonderful thing, but a look at country growth rates over recent centuries and today reveals, with striking consistency, decreasing returns at the level of national economies. As long as that remains true, convergence is a simple necessity for comeattibleness.

What to do? As I suggested, nothing here speaks to whether redistribution measures, growth policies, or other strategies are best. Yet I would not dismiss the effectiveness of at least some rich-to-poor transfers. Thanks in no small part to the transfer of medical goods, finance and technology, even the world’s poorest people have seen dramatic growth in child survival and life expectancy – a prima facie gain in ability to ‘come at’ life’s necessaries. Internationally supported education, for its part, has yielded very large gains in female literacy – another example of why we should not despair of our ability to help.

Finally, I believe that I do have obligations to people in have-not countries. Every day I consume things produced through a worldwide value chain that includes many such countries. Meanwhile, my consumption pours greenhouse gasses into the atmosphere – a huge, life-and-death peril for the world’s poorest. That does not put me on the hook for solving inequality and all its effects, but neither does it leave me free to ignore them.

FL: I will respond in two parts. GDP is a relatively objective measure. Anything else takes us into the realm of value or moral judgements. How do we define happiness, let alone measure it?

Next, the marginal utility of income may decline if utility does not depend on one’s relative income, but only on one’s absolute income. As my thesis adviser pointed out many years ago, relative income matters – for better or for worse. Indeed, the world’s wealthiest individuals look forward to the annual rankings of the wealthiest people in the world in order to see where they stand in relation to their peers. They can never have too much. Consequently, for them, and for most people, marginal utility of income does not decline as their incomes increase.

Even if this were not the case, redistribution – especially across borders – entails costs: the costs of setting up and administering the system. These costs represent a waste of resources. Furthermore, while some of the wealthiest may voluntarily choose to transfer some of their wealth to others (consider the Giving Pledge movement started by Gates and Buffett), it is unlikely that there would be any support within any of the ‘wealthy’ countries for an involuntary regime of redistribution.

In respect of the data for Canada, would Canadians have been better off with lower growth rates and a more equitable distribution of the gains? Did inequality negatively impact growth in Canada or elsewhere? There is no evidence to suggest that this has been the case. China, which has even greater income inequality (and inequality of opportunity) than Canada, has experienced very strong growth over the past two decades. Would the country and its people have been better off with ‘Mao’ economic outcomes – slow growth and greater equality? I doubt it.

I now turn to shorter responses on additional points. The evidence does not support the contention that there are decreasing returns. Productivity growth does fluctuate, and tends to be more rapid when a country transitions from an agrarian society to an industrialized one. However, there has not been any consistent downward trend in productivity growth rates among the wealthy countries during the past century.

While I can agree that there may be some merit in promoting health and education initiatives in the poorest countries, the key question is: how to do this? Money can easily be wasted or worse, stolen, as has been the frequent case in the past. And redistribution does impact incentives and growth in the wealthy countries. Perhaps the poorer countries would be better off if they were more fully integrated into the global economy, and if corruption were attacked more forcefully.

As for female literacy and education: do we really wish to get into a debate about the role of religion and culture?

JR: I now worry that we are covering so much ground so quickly that we may be talking past each other. This may in part be about differing scales and time horizons. The big growth out of poverty in Japan, East Asia and China involved gains of around eight to 10 percent per year. Is that a plausible rate for rich countries, the growth of which over the past century has been more in the two to three percent range? If not, then we do need more catch-up – that is, convergence – to achieve basic accesses to life’s “necessaries,” not to mention conveniencies. For what I called the striking consistency in the relevant convex production curves on this scale, see the wonderful empirical work of economic historian Robert Allen. I am not trying to say more than that about diminishing returns.

Similarly, while I agree that relative income matters, an added $200 a year is not going to change the ranking of anyone in a list of the world’s wealthiest. In the hands of new parents on the Bangladesh delta, however, that $200 may save a baby’s life. There is no plausible equivalence between the value of $200 to Gates or Buffet (whose Giving Pledge I applaud) and its value to those parents. Our friend Adam Smith addressed both relative income and (what we call) decreasing marginal utility in his own 18th century context. Similar observations, grounded in his “necessaries and conveniencies” framework, led him to favour taxing the rich.

The early March issue of The Economist included a summary article about recent research on inequality and GDP growth. There is indeed evidence associating inequality with slow GDP growth, but since we do not really understand what causes and sustains GDP growth I would take a cautious view about such evidence. I find the question that you pose before you raise this evidentiary point much easier. If Canada’s GDP over the generation that I cited had grown by, say, two-thirds of its actual gain, with the (still very large) aggregate benefit distributed equally to all Canadians, then I would certainly consider us better off: working class parents could expect better prospects for their children rather than drudgery with no gain; the poor would see – for them, hugely significant – gains as opposed to wrenching losses; while the wealthy would still be advancing apace with each other.

This assessment brings me to your opening paragraph about definitions, measurement and judgement. Your question about Canada was easier for me to manage than The Economist’s question about GDP growth. This was so for a basic reason: even without a precise measure of what makes the necessaries and conveniencies of life easily come at, the hypothetical Canadian scenario seems to me clearly superior on that score to what actually happened. Happiness – an alternative standard, and not my choice – is indeed hard to measure. That has not stopped smart people like Richard Easterlin from trying, with interesting results that are no less solid than the GDP-inequality literature that The Economist cited.

But when we choose concepts and measures to evaluate economic performance, we cannot avoid value judgements. Adam Smith was a moral philosopher. If we choose per-capita GDP, we are choosing, contra Smith, to value Bill Gates’ marginal $200 as equivalent to that of a Bangladeshi family. If we choose the logarithm of per-capita GDP (as the Human Development Index does, as a rather standard, rough-and-ready quantification of diminishing utility), then we give more weight to that baby’s chance to live. En passant, why not choose as our measure, say, the percentage of children surviving to maturity? That is a far better proxy for the necessaries of life than GDP, with the advantage that it is easier to measure and no less ‘objective.’ North America comes out as no star on this metric.

There is no single best measuring tool. I for one would include not just longevity, but also health, capabilities and opportunity. But a lens that is blind to our children’s survival is itself impoverished.

FL: Indeed – let us not spend more time on definitions and immeasurables. We can debate them endlessly. Yes, it would be wonderful if a rising tide raised all ships. Yes, it would be wonderful if we could achieve higher rates of growth without destroying our environment – and that, in the process, the greatest beneficiaries would be those at the lowest end of the income spectrum, both across countries and within each country. But the real question is: how can this be done, if at all?

How would you propose redistributing income across countries and within countries? How much should be redistributed? Who would be responsible? How much would it cost to manage this redistribution? What would be the economic effects of this grand scheme? And could you ever get people to voluntarily agree to this grand scheme?

To say that redistribution must be done regardless seems an elitist and paternalistic argument. In other words, those who favour massive redistribution know better – and indeed have higher moral standards – than others, and therefore should be allowed to implement their views, with details to follow. Of course, these details have not really been worked out or even considered.

JR: I propose no grand scheme. For the reasons you suggest and more, it would be unworkable. Instead, let one hundred – actually, many more – redistributive flowers bloom. There is already significant redistribution occurring – domestically and internationally. We can build on this if we pay attention to the effects of what we are doing, and try new things accordingly. We can wind down what does not work, ramp up what does, and continue to experiment. Individuals, corporations, civic groups, governments, international organizations, researchers – all can initiate or join such experiments.

On the international scene, Canada is already pioneering this type of approach – in cooperation with the Gates Foundation, to pick up an earlier thread in our discussion – by sponsoring Grand Challenges Canada (GCC). GCC supports promising innovations engaging the public, private, academic and voluntary sectors to tackle key challenges in global health. I cannot promise the wonderful outcome that you outline in your second volley, but I am certain that we can get much closer to it than we are today.

Nor can we say in advance what all the effects will be – one reason for which we need experiments, not grand schemes, if we wish to attack grand challenges. But we do have empirical data on which to draw. The recent IMF paper Redistribution, Inequality, and Growth, by Ostry, Berg and Tsangarides, analyzes the fullest relevant data set of which I am aware, and harvests lessons about how redistribution has been – but is not always – ‘benign’ in its impact on GDP. The world has not maxed out on redistribution. If we have the will, then we certainly have the room for new experiments.

Political will to redistribute need not be, and should not be, elitist. To the extent that governments are involved, decisions should be made democratically: this entails working to improve democracy in rich and poor countries alike. Consider, for example, one of the largest redistributive programmes now operating: the PEPFAR initiative in the US. George W. Bush proposed PEPFAR, a large-scale initiate to prevent and treat AIDS in have-not countries – in no small measure because of pressure from one of his key constituencies: conservative Christians. That constituency is generally considered the very opposite of elitist – perhaps because they tussle with biologists over evolution. Here, however, they worked with biomedical scientists, leftist activists, major corporations, philanthropists – including, again, Gates – and ordinary people to put tens of billions of public dollars into saving lives in poor countries. Against tough odds, the effort has been a huge success and retains broad political support. So let us expand this initiative: defeating the AIDS epidemic is now in our grasp.

Let me conclude where I began: Adam Smith first introduced the ‘invisible hand’ not in his oft-quoted masterpiece The Wealth of Nations, but in his earlier one, The Theory of Moral Sentiments. That book analyzes how our emotional constitution plays out in human society to yield beneficial outcomes that we could not achieve with the poor power of our intellects. Smith argues that this is because intellects cannot calculate consequences well out into the future. Using the language of our own discussion here, we might say that Smith’s opening point is that the utility of both rich and ordinary people depends not only on their own absolute and relative wealth, but also on the fate of others. Smith said it better: “How selfish soever man may be supposed, there are evidently some principles in his nature, which interest him in the fortune of others, and render their happiness necessary to him, though he derives nothing from it except the pleasure of seeing it.”

Let us build on that pleasure.

FL: Let me make, by way of conclusion, three sets of observations – all of which are critical for understanding the essence of this debate. First, economic growth has likely been the most important driver for bringing large numbers of people out of poverty: consider the Industrial Revolution in the late 1700s and 1800s; China and select Southeast Asian or ASEAN countries over the past 25 years; and the US during the 1990s. Arthur Okun once argued in favour of policies aimed at maintaining high rates of growth. During the 1990s in the US, such policies were pursued, pushing unemployment rates even below 4 percent for some periods by the end of this decade, and reducing the incidence of poverty in the country by the largest amount – in absolute and relative terms – seen in the past 50 to 60 years. Income redistribution policies and social welfare programmes cannot compete with economic growth to tackle poverty. For this reason, I take issue with all of the economists who promote austerity and balanced budgets as the solution to our economic problems.

Second, according to World Bank data, there are about a dozen countries in which 70 percent or more of the people live on less than $2 per day. I, like many in the West, spend more than this daily on coffee. However, these countries by and large share the following unpleasant characteristics: they tend to be among the most corrupt and least democratic in the world. Putting more money in the hands of the governments running these countries will do little to alleviate poverty, but will increase the offshore bank accounts of the kleptocrats who govern.

Finally, the following 29 countries all have lower Gini coefficients – hence, supposedly, more equal distribution of incomes – than Canada: Afghanistan, Australia, Austria, Bangladesh, Belarus, Bulgaria, Czech Republic, Denmark, Egypt, Ethiopia, Finland, Germany, Hungary, Iraq, Kazakhstan, Luxembourg, Montenegro, Netherlands, Norway, Pakistan, Romania, Serbia, Slovakia, South Korea, Slovenia, Sweden, Tajikistan, Timor Leste and Ukraine. But it is not all obvious that greater income equality produces more desirable economic outcomes. Many of these countries have very high levels of poverty, as measured by the proportion of the population living on less than US$2 per day.

Furthermore, only five of the above 29 countries with greater income equality have a higher GDP per capita than Canada. Thus, income inequality does not necessarily translate into lower GDP per capita – that is, lower average standards of living. In addition, greater income equality does not necessarily translate into greater civil rights or less corruption (Canada ranked 9th out of 177 countries in Transparency International’s rankings).

James Radner is Assistant Professor in the School of Public Policy and Governance, University of Toronto, where he also directs the Boreal Institute for Civil Society.

Fred Lazar is Associate Professor of Economics in the Schulich School of Business, York University (Toronto).

(Photograph: The Canadian Press / Shehzad Noorani)