Sunday, February 5, 2017

In Search of the Elusive Victims of Globlization

The 2016 US Presidential election has placed trade policy high on the national agenda. Both Bernie Sanders, on the left, and Donald Trump, on the right, campaigned on overtly protectionist platforms. Now that Trump is in office, he has begun implementing his program of “buy American, hire American.”

In response, many members of the economics profession, always a bastion of free-trade sentiment, have taken a new look at something they always knew but did not always like to talk about: the fact that trade creates winners and losers. In a widely cited paper, “The China Shock,” David Autor and colleagues show that the losses from trade shocks to the US economy are larger and more persistent than many had thought. Such research makes it understandable how politicians can assemble victims of trade shocks into winning coalitions.

Although Trump and Sanders have directed most of their critique of global trade at the way it creates losers in the US economy, other critics are more concerned with the effects on US trade partners. Taking advantage of the media attention drawn by their sometimes disorderly protests against the Seattle meetings of the World Trade Organization in 1999, these critics emphasize that trade creates victims in poor countries as well as rich.

In an entry, “Defining the Anti-Globalization Movement,” in the Encyclopedia of Activism and Social Justice, Mark Engler points out that there is no unified movement that fits that term. Instead, there is an informal coalition of trade unionists, environmentalists, indigenous rights activists, organizations promoting sustainable development, and anti-sweatshop campaigners.  Many of these groups spurn the “anti-globalist” label, preferring terms like “global justice movement” or even (standing the term on its head) the “globalization movement.”

What unites these groups, Engler says, is the belief that the corporate-led globalization of the past quarter century has exacerbated global poverty and increased inequality. Many of them frame their mission as one of opposition to neoliberalism, by which they mean policies including privatization of public industries, opening markets to foreign investment and competition, removing controls on capital flows, reducing tariffs and other trade barriers, and ending government protections for local industry. Movement participants, Engler says, “argue that these policies have created sweatshop working conditions in the developing world, threatened unionized jobs and environmental protections in the global North, benefited the wealthy at the expense of the poor, and endangered indigenous cultures.”

At the risk of oversimplification, I think there is a unifying theme here, which I will call the globalization hypothesis : The proposition that free trade (or neoliberalism, if you prefer) has enriched elites at the expense of the vulnerable in both the developed and the developing world.

The search for evidence

At the same time trade issues were making their way regularly into the headlines, I was working on an unrelated project using some large data sets that were rich in global economic, social, and political indicators. It occurred to me that it might be possible to get some insight into the validity of the globalization hypothesis by looking for critical differences in these indicators between countries that were more open to trade and those that were less open.

One of the data sets is the Economic Freedom Index from the Fraser Institute. It includes a component, Freedom to Trade Internationally, that seems to capture important parts of the neoliberal agenda, as described by critics like Engler. Fraser’s trade freedom score, as I will call it, is calculated on a scale of zero (least free) to ten (most free), based on indicators for tariffs, regulatory trade barriers, compliance costs for importing and exporting, black market exchange rates, controls of the movement of capital and people, restrictions on foreign ownership and foreign investment, capital controls, and freedom of travel.

The other data set I will use here is the Legatum Prosperity Index . Its creator, the Legatum Institute, is more sympathetic to the global justice perspective, in that it views prosperity as the creation of a better life for individual people rather than as just the accumulation of material wealth. The Legatum index organizes the data into distinct economic and noneconomic components, or “pillars”: Economic quality (macroeconomic indicators, and financial institutions); business environment; governance (political participation, transparency, and rule of law); education; health; safety and security (both national security and personal safety); personal freedom (human rights, legal rights, and tolerance); social capital (personal relationships, social network support, civic participation); and natural environment.

The Fraser Institute’s data set covers 159 countries while that from the Legatum Institute covers 149. Both indexes are available for an overlapping set of 144 countries.


First, let’s check to see if these data support the proposition that free trade is associated with strong macroeconomic performance and a good business environment. Neoliberals certainly think that it does, and I think many global justice advocates do not dispute it at the macro level. Here is a scatter plot of the Fraser trade freedom scores against the Legatum economy and business environment scores:

As expected, there is a solid positive relationship in both cases. The correlation coefficient for the economy is 0.63 and for the business environment 0.71, where a coefficient of 1.0 would indicate a perfect fit. Both coefficients are statistically significant at the 0.01 level. No surprises here.

But what about the noneconomic effects of free trade? That is where we might expect evidence relevant to the globalist hypothesis to show up. To see if it does, I first calculated a “social prosperity” score for each country, consisting of the combined scores, scaled zero to 100, on the Legatum components for governance, education, health, security, personal freedom, social capital, and environment. Here is the result:

This result is surprisingly similar to the chart for the economic and business components: The correlation coefficient between social prosperity and freedom to trade is 0.68, again positive and statistically significant. That is also true for each of the individual components of social prosperity. Putting correlations in parentheses, in countries that are open to free trade, people tend to be better-governed (0.67), better educated (0.61), healthier (0.60), more secure (0.64), and freer (0.57). They also enjoy more social capital (0.43) and a better natural environment (0.43).

There is still a reason to be skeptical, however. Maybe what is really happening is that a few wealthy, free-trading countries are pulling up the average on the right side of the chart at the expense of their poorer third-world trading partners. If that is the case, we would expect the relationship between trade and social prosperity within the developing group to be negative, or at least much weaker than within the entire sample.

We can check that possibility by dividing the sample into two parts, one for countries that are already wealthy and the other for the developing world. The next chart does that. The blue diamonds represent the 35 members of the Organization for Cooperation and Development; the red squares are the remaining 109 non-OECD countries.

The correlation between trade freedom and social prosperity remains positive (0.63) and statistically significant within the non-OECD group, just as it is for the group as a whole. There are some outliers, however.

At the extreme left of the chart, three countries, Iran, Venezuela, and Argentina, are the least open to trade of any countries in the world. All three lie above the trend line, indicating that they have better-than-expected social prosperity scores. They are not exactly poster-children for social prosperity — Iran  and Venezuela are below the world social prosperity median, and Argentina is only a little above it — but they do suggest that at least a few countries have closed themselves off to trade without suffering disastrous social consequences. The data don’t give much of a clue as to what the three might have in common, except that all of them have poor scores on governance and somewhat better than expected scores on education.

Meanwhile, at the extreme right, we find Singapore and Hong Kong, the two countries that are most open to trade of all. Both of them also have higher than expected social prosperity scores. Singapore ranks number one in the world in safety and security and number two in health, although personal freedom, where it ranks ninety-seventh, is clearly a weak spot. Hong Kong has excellent scores for health, safety, and education, but it ranks a smoggy ninety-eighth in terms of its natural environment.

Interestingly, there is no significant correlation between freedom to trade and social prosperity within the OECD. Perhaps that should not come as a surprise, since the OECD is not a randomly selected group. Countries are not invited to join unless they have open economies, so there is not enough variation among members in that regard to give statistically meaningful results.

What does distinguish OECD countries even more than trade freedom is good governance — a Legatum category that includes participation, transparency, democracy, and rule of law. Eighteen of the twenty highest governance scores in the world belong to OECD countries. (The only non-OECD countries in the group are number 18, Singapore, and number 19, Uruguay.) Since good governance correlates highly with other elements of social prosperity, it is not surprising that the OECD countries lie in the upper-right-hand corner of the chart.

The issue of equality

Even when we split the sample into OECD and other countries, we have not really come to grips with one of the central concerns of the global justice movement. Its adherents would rightly point out that even if the national averages for social prosperity indicators might look good for countries that trade, the individual elements of prosperity, whether health care, education, or personal security might still be unequally distributed within each country. If so, then our charts, which show only country averages, might mask a tendency for trade to benefit elites within each country at the expense of the rest of society.

As a starting point, we can look at the relationship between openness to trade and equality of income distribution within countries, using data from the UN Human Development Report. That report includes Gini indexes of income inequality for 116 countries that overlap with the Fraser trade freedom data. Gini indexes range from zero (completely equal distribution) to 100 (maximum inequality). The most equal distribution in the UN data set is Sweden, with an index of 25, followed closely by Norway. The least equal is Namibia, at 63.9, with South Africa a close second. The United States has the second-highest score in the OECD, surpassed in inequality among that group only by Mexico. The US Gini index of 40.8 is just above the global median of 39.9.

Across the whole sample of countries, the correlation coefficient of the Gini index with the trade freedom score is -0.31, not a tight fit but statistically significant at a 0.01 level of confidence. Since the Gini index increases as equality decreases, the negative correlation coefficient means that the countries that are most open to trade tend to have more equal income distributions. However, that correlation turns out to be attributable entirely to the association between the two variables within the OECD. The correlation coefficient within the non-OECD group is -0.02, negative but not statistically different from zero. There is, then, no evidence here to suggest that trade freedom is associated either in one way or the other with pure inequality of income distribution within our group of developing countries.

However, it is the relationship between trade and inequality for the noneconomic elements of social prosperity that really goes to the heart of the globalization hypothesis.  Anecdotally, we know that there are countries where elites are well educated and live long lives while the poor are illiterate and die young. Is there any evidence in our data set that there is a general tendency for countries that are open to trade to fit that pattern?

Unfortunately, we don’t have explicit data on the within-country distribution of most of the social prosperity indicators. What we do know, however, is that the people who put together the Legatum index were aware of this issue and took steps to deal with it. They did so by including indicators within each component of their index that minimize the likelihood that countries with high degrees of social inequality could achieve high scores. For example:
  • The Legatum education component includes adult and youth literacy rate, a Gini index for education equality, primary school completion rate, secondary school enrollment, and the ratio of school enrollment for girls and boys. Those indicators are weighted to account for about half of the education score. As a result, education indicators that might be skewed toward national elites, such as test scores and enrollment in top universities, cannot by themselves give a top score to a country as a whole.
  • The health component includes access to basic sanitation, immunization rates, deaths from tuberculosis, and life expectancy at birth. The presence of such indicators would pull down the score of any country in which only elites live healthy lives.
  • The personal safety and security component includes measures of malnutrition, homelessness, deaths from civil and international conflicts, number of refugees, and safety of walking at night. These indicators would generate low scores in countries where only elites were safe and secure.
Other components include similar indicators measuring items of social prosperity that are by their nature beneficial to all, not just to elites. That is true even of the components relating to economic quality and business environment. You can find the full listing of indicators for each component in an appendix to Legatum’s Methodology Report.


The findings reported here do not constitute an actual refutation of the globalization hypothesis, as we have formulated it — the proposition that free trade systematically enriches elites at the expense of the vulnerable in both the developed and the developing world. What they do show is that once we move away from case studies and anecdotal evidence, broad statistical evidence to support the hypothesis is hard to find. On the contrary, if anything,  there seems to be a significant tendency for people to live healthier, more secure, and freer lives in countries that are open to global trade than in those at similar levels of development that are less open.

I doubt if that finding is really a surprise to thoughtful supporters of the global justice movement. Even in a country where trade brings gains in broad indicators of social prosperity, such as infant mortality and girls’ literacy, there can be individual losers, and the gains, even among the winners, may not be distributed as widely as one might like. The real issue facing the global justice movement is what to do about it.
The positive association between trade and social prosperity at the national level suggests that the proper response, in both wealthy and developing countries, should take the form of policies that spread the gains of trade widely and help individuals to adjust to change, rather than in rolling back the process of globalization itself.

In an essay on globalization in The Scientific American, UC-Berkeley economist Pranab Bardhan puts it this way:
There is no race to the bottom in which countries must abandon social programs to keep up economically; in fact, social and economic goals can be mutually supportive. Land reform, expansion of credit and services for small producers, retraining and income support for displaced workers, public-works programs for the unemployed, and provision of basic education and health can enhance the productivity of workers and farmers and thereby contribute to a country’s global competitiveness. Such programs may require a rethinking of budget priorities in those nations and a more accountable political and administrative framework, but the obstacles are largely domestic.
Conversely, closing the economy to international trade does not reduce the power of the relevant vested interests: landlords, politicians and bureaucrats, and the rich who enjoy government subsidies. Thus, globalization is not the main cause of developing countries’ problems, contrary to the claim of critics of globalization — just as globalization is often not the main solution to these problems, contrary to the claim of overenthusiastic free traders.
Our data support Bardham’s view, inasmuch as the correlation of good governance with other indicators of social prosperity is even stronger than that of openness to trade. Ultimately, each country is responsible for its own governance, but wealthy countries that trade with nations that are poor and badly governed can support local efforts. They can insist that their own companies not just comply with local labor and environment standards, but aim higher than those local standards. They can insist that their own nationals do not corruptly conspire with local economic and political elites at the expense of the poor. And they can politely but firmly resist the suggestion of any of their own citizens that the best way to help people on lower rungs of the global economic ladder is to refuse to buy anything from them.

Originally posted at Fabius Maximus

No comments:

Post a Comment