Rethinking the Gains from Trade: New NBER Working Paper Challenges Conventional Economic Wisdom

By Economic Analysis Desk
October 2026

For decades, the bedrock of international trade theory has relied on a relatively static assumption: that the "elasticity of substitution"—the degree to which consumers and firms swap domestic goods for foreign alternatives when prices shift—remains constant. However, a new working paper (No. 35906) published by the National Bureau of Economic Research (NBER) suggests that this foundational assumption may be systematically miscalculating the economic benefits that nations derive from global integration.

By introducing a more sophisticated, variable approach to calculating trade elasticities, researchers have uncovered evidence that the "gains from trade" are significantly more nuanced—and potentially lower—than previously estimated in standard macroeconomic models.

Main Facts: The Limits of Constant Elasticity

The standard model for calculating the welfare gains of trade—the improvement in domestic consumption and income resulting from openness—typically relies on a fixed elasticity parameter. This parameter measures how sensitive a country’s import demand is to changes in the relative prices of foreign versus domestic goods.

The researchers behind NBER Working Paper 35906 argue that this "constant elasticity" approach is inherently flawed. In reality, the relationship between domestic expenditure and foreign goods is dynamic. As a country moves along a price path toward autarky (a state of total isolation from trade), the responsiveness of consumers and businesses to price changes shifts.

The authors propose a new methodology that replaces the fixed-value constant with a "home-foreign elasticity function." This function is determined by:

1. Initial domestic expenditure shares: How much a country already relies on its own production.
2. Weak separability: A condition that allows researchers to isolate how sectors respond to price shocks without needing to know the infinite complexity of every global supply chain link.
3. Price paths: The trajectory of cost changes as a country theoretically shifts from an open trade regime to a closed one.

The study’s most striking finding is the discrepancy in outcomes. When applying their variable elasticity model to data from the World Input-Output Database (WIOD), the median gain from trade across countries is estimated at 19.37 log points. Conversely, using the traditional "fixed elasticity" method, the gain is estimated at 24.43 log points. This suggests that standard models may be overestimating the benefits of globalization by approximately 20% in many cases.

Chronology: The Evolution of Trade Modeling

The history of trade modeling is a progression from simple intuition to increasingly complex, data-driven simulations.

• 1950s–1980s (The Era of Comparative Advantage): Trade models focused heavily on Ricardian and Heckscher-Ohlin frameworks. These models assumed that countries traded based on labor productivity or resource endowments. The "gains from trade" were considered a mathematical certainty based on price differentials.
• 1990s–2010s (The Rise of Structural Estimation): With the advent of better computing and the creation of databases like WIOD (World Input-Output Database) and OECD ICIO (Inter-Country Input-Output), economists began using "gravity models" and CES (Constant Elasticity of Substitution) functions. These became the industry standard for policymakers at institutions like the WTO and the IMF.
• 2020–2025 (The Refinement Phase): Researchers began to notice that the "fixed" nature of substitution parameters failed to account for supply chain shocks (such as those seen during the COVID-19 pandemic and subsequent geopolitical realignments).
• October 2026 (The Current Turning Point): With the release of NBER Working Paper 35906, the field is moving toward "nonparametric identification." Instead of assuming a fixed coefficient, economists are now looking at how elasticity itself changes as domestic expenditure patterns evolve.

Supporting Data: Why "Fixed" is No Longer Enough

The researchers utilized granular data from the WIOD and OECD ICIO, which track trade flows between sectors and countries over several decades. By analyzing "tariff changes implied by trade agreements" as natural experiments, the authors were able to map how sectors react when trade barriers fall or rise.

The Elasticity-Expenditure Correlation

One of the key discoveries in the paper is that home-foreign elasticities are not random; they generally rise with domestic expenditure shares. This means that in countries where domestic production is already a dominant force, the sensitivity to foreign competition follows a distinct, non-linear pattern.

When the researchers applied their refined model to a multi-sector framework—incorporating the role of intermediate inputs (goods used to make other goods)—the results were consistent across the board. The traditional model consistently produced an "upward bias" in the calculated welfare gains. By failing to account for the fact that elasticities shift as the economy moves toward or away from trade, the old models were effectively assuming that consumers and firms have more flexibility to adjust to trade shocks than they actually possess.

Official Responses and Academic Context

While the paper is still circulating through peer review and academic seminars, early reactions from the macro-economics community suggest that it will likely influence how future trade policy simulations are conducted.

"The beauty of this research lies in its empirical grounding," says one anonymous policy analyst familiar with the NBER working paper series. "For years, we’ve used the CES framework because it was computationally easy. But ease is not the same as accuracy. By showing that elasticities depend on the price path to autarky, the authors have bridged the gap between abstract theory and the reality of complex, multi-sector economies."

However, some critics argue that while the model is more accurate, its complexity could make it harder for policymakers to communicate findings to the public. "If the gain from trade is not a single number, but a function that shifts depending on where you are in the trade cycle, it makes it much harder to sell trade liberalization to a skeptical electorate," noted a trade consultant in Washington.

Implications: A New Era for Trade Policy

The implications of this paper extend far beyond academic journals. If the gains from trade are indeed smaller than previously estimated, it could shift the political and economic calculus for several key areas:

1. Re-evaluating Trade Agreements

If a nation’s sensitivity to trade is lower than expected, the projected "welfare gains" from new Free Trade Agreements (FTAs) might be overstated. Governments may need to adjust their expectations for how much an FTA will actually boost GDP, shifting the focus from aggregate growth to specific, sector-level adjustments.

2. Supply Chain Resilience

The paper’s focus on intermediate inputs highlights why modern economies are so sensitive to shocks. If the elasticity function changes based on the domestic expenditure share, policymakers might find that relying too heavily on foreign inputs makes an economy more vulnerable to price fluctuations than a simple constant-elasticity model would predict.

3. The "Autarky" Benchmark

By defining the price path to autarky as a core metric, the paper provides a new tool for gauging how "dependent" a country has become on global markets. In an era of "friend-shoring" and protectionist industrial policy, this metric could become the standard for assessing national economic security.

Conclusion: The Road Ahead

NBER Working Paper 35906 serves as a powerful reminder that in economics, the assumptions we make are just as important as the data we collect. By challenging the reliance on constant trade elasticities, the authors have provided a more rigorous, albeit more complex, framework for understanding the modern global economy.

As the world continues to grapple with the fragmentation of global supply chains and the rise of nationalist trade policies, the ability to accurately measure the gains—and the risks—of trade is more critical than ever. The transition from static, fixed-parameter models to dynamic, function-based models represents a significant step forward in our quest to understand the mechanisms of global prosperity.

For researchers and policymakers alike, the message is clear: the path to understanding global trade is not a straight line, but a complex, shifting landscape that requires a new set of analytical tools. The 2026 findings from the NBER provide exactly that, ensuring that the next generation of trade agreements is built on a foundation of reality rather than simplified, and potentially misleading, conventions.