Skip to main content

Trade liberalization: Poverty’s friend or foe?

 

They are the familiar images of global trade talks: streets turned into battle zones, conference centres into fortresses. Entrenched positions on both sides of the barricades have effectively stifled public debate on trade liberalization. Is the developing world healthier, wealthier, better fed, and better educated thanks to trade-led economic growth? Or has it triggered a rising economic tide that will float yachts but threatens to swamp simpler craft anchored in poverty? There is precious little quantitative research to support either position.

“We know that trade liberalization is supposed to be good for growth,” says Dr John Cockburn of Université Laval in Québec City, Canada, “but we’re not quite clear on who wins and loses, or how freer trade affects poverty.”

One of the main stumbling blocks is the lack of an effective means for linking changes in trade policy to changes within individual households. This connection is critical to estimating the impact of trade on poverty.

To date, economists have used tools, such as computable general equilibrium (CGE) models, to gauge the effects of macroeconomic shocks — trade liberalization or tax reform, for instance — on the economy as a whole. Most models, however, do not or cannot make the link to the effects on individual households. [See: Computable General Equilibrium Models: A primer]

“CGE models generally work with categories of households, such as rural farm and rural non-farm households,” says Cockburn. “They do not distinguish between the poor and non-poor within these categories, nor do they give any idea of the proportion of each in any of the household categories.”

To overcome this hurdle, Cockburn drew on the work of a team of developing-country researchers from the Poverty and Economic Policy (PEP) Research Network, which he co-directs. The network is part of the International Development Research Centre’s (IDRC) Micro Impacts of Macroeconomic and Adjustment Policies (MIMAP) program initiative. MIMAP was launched in 1990 to help developing countries measure and analyze poverty and to design improved policies and poverty alleviation programs.

The PEP researchers developed CGE models of their respective African and Asian countries to forecast how changes in trade policy, such as tariff reductions, affect household income and the consumer prices they face. By combining the predicted income and consumer price results of the models with data from standard household surveys, the researchers developed an accurate picture of the real income distribution effects within each category of households. The researchers then applied standard poverty indicators to this data to develop a poverty profile before and after trade liberalization policies were introduced.

In this case, the indicators were: the poverty headcount to measure the percentage of households under the poverty line, the poverty gap to assess how far below the poverty line households fell, and poverty severity to examine the distribution of income among poor households. [SeeA sample outcome]

Surprising results

The model was piloted in eight countries: Bangladesh, Benin, India, Nepal, Pakistan, the Philippines, Senegal, and Vietnam. Each research team analyzed the impact on poverty of trade liberalization policies since 1995.

The PEP research team is now analyzing the data from the national studies to see if they can draw any broad conclusions about the links between trade liberalization and poverty. The analysis, he cautions, is still in its preliminary stages, but the results have already produced a few surprises.

For example, conventional wisdom suggests that the heavily protected sectors of the economy — generally the industrial sector in developing countries — would shrink as trade liberalization removes tariffs and national markets open to competition from imported goods. Other sectors — commonly agriculture — would expand to exploit new opportunities afforded in the export market.

“Prior to our study,” says Cockburn, “we thought agriculture would be the big winner as a result of trade liberalization.”

“But,” says Cockburn, “the ‘export push’ that comes from trade liberalization doesn’t benefit the agricultural sectors as much as the industrial sectors.”

The reason, he argues, is three-fold:

  • The difference in tariff levels between the agricultural and industrial sectors was not as great in 1995 as it was in the heyday of industrial protectionism in the 1970s and ’80s, so the shock of lowering trade barriers was not as pronounced as it might have been; and,
  • Local industries benefited enormously from cheaper imported inputs, effectively offsetting the drop in domestic prices caused by competition with lower cost foreign manufactured goods.
  • Agricultural sectors are less export-oriented than industrial sectors in most of these countries, and thus benefit less from trade liberalization.

Wage gains

The researchers also examined the effect of trade liberalization on wages and income. Their analysis did show a relative rise in wages and incomes but not for the reasons expected. It was generally assumed that the lack of tariff protection in the agriculture sector would better position labour intensive agricultural industries to seize opportunities in export markets after liberalization. This, in turn, would increase the demand for unskilled labour and push wage levels upward. The study clearly shows that agriculture did not benefit as expected from freer trade. So what accounts for the relative rise in wages and incomes? It was, in part, the buffering effect of the service sector, says Cockburn.

“Forty to 60% of labour [employment] comes from the service sector. Because the service sector is not directly affected by freer trade, it did not face import competition in any significant way and its price levels are relatively unaffected. As a result, wages remained relatively stable and, relative to other factors such as the drop in consumer prices, they increased. So, in the economy as a whole, the service sector has a stabilizing effect on workers wages. This is especially important for the urban poor who make up most of the workers in this sector.”

Questioning conventional thinking

Cockburn’s analysis challenges many long-held notions about trade liberalization and poverty.

“We went into this thinking that the effects of freer trade on poverty were going to be country specific. But as we work with the data, there seem to be some rules that apply to all the countries. I can say that, generally, trade liberalization has positive effects on welfare and poverty, especially for urban households.”

Policy links

Despite the generally positive effects of freer trade, many households still face difficulties. Indeed, one of the clear lessons from the PEP study is that governments need to pay close attention to how workers, especially agricultural ones, and their families adjust to changes brought on by freer trade.

“To adjust successfully you have to really look at the export possibilities that open up rather than simply trying to compete with foreign firms on the domestic market,” says Cockburn. “You need to know where big changes will occur, be able to remove any obstacles to adjustment, and lessen the negative impact of that change. It’s a juggling act.”

The juggling act is made more difficult by shrinking government revenues as income from tariffs drop. “How much government revenues will drop is a matter of some debate,” says Cockburn, “but there is a general agreement that they will fall because of trade liberalization.”

This sobering fact leaves policymakers with two choices: cut spending or find other sources of revenue, typically through taxes. Cockburn and the PEP researchers are now “modeling” various options to see which ones have the most positive poverty and welfare effects.

Understanding the trade-offs in the choices they make, should help policymakers better tailor solutions to fit their economies. Will it ensure more equitable growth from trade liberalization? We can all look to the barricades for the answer.

A sample outcome#sample

“If our model shows a 5% increase in the average income of one category of households, we could apply this increase to survey data and see how the income of each household within the category changed,” says Cockburn. “By comparing income distribution before and after trade liberalization to an established poverty line, we can see how many households were lifted out of poverty.”

Computable general equilibrium models: A primer #computable

Standard CGE models break an economy down into five main components — producers, factors of production (workers, land, etc.), households, government, and the rest of the world — and then develop equations that describe the behaviour of and interaction between the various components. By changing variables, researchers can develop “what if ” scenarios to assess the outcome of proposed policy changes and to identify the important channels through which the changes act.

But any model is only as good as the data available to put in it and the assumptions on which it is based. While the use of standardized household surveys helped to ensure the reliability of input data, assumptions were tested for robustness.

In the PEP model, for example, researchers made assumptions about how local producers would adapt from selling locally to selling in export markets. Researchers test their assumptions and adjust their models by replicating a known event, for example the effects of earlier tax reform. Nevertheless, models do not replicate the exact functioning of any economy. Good models are in Cockburn’s words “thinking tools.

Kevin Conway is a senior writer in IDRC’s Communications Division.