The International Copper Cartel, 1935–1939: the good cartel?

Abstract

A legacy of the pre-Great Depression period was the development of new copper capacity outside the USA. In order to avoid the utilization of this excess capacity in a competitive and possibly pyrrhic way, on March 28, 1935, the International Copper Cartel was formed. While production of cartel members was limited by the use of quotas in times of lower demand, during booms, the cartel removed all quotas in order to restrict the entry of the competitive fringe into the industry. In fact, many authors have considered the latter as a return to non-cooperative conduct. An alternative view, introduced by Montero and Guzmán (J Ind Econ 58(1):106–126, 2010), is that cooperation among members in booms could have been related to output-expanding strategies with the purpose of restricting entry of the competitive fringe in the industry through their own output expansion.

In this paper, a theoretical model is developed in order to differentiate between cartel behavior in the presence of a competitive fringe during booms when firms are not cooperating. It shows that the price elasticity of the cartel supply should be smaller in booms than during recessions, when a non-cooperative behavior is considered in booms, while this may or may not be the case if firms continue cooperating—through an output-expanding cooperation—during booms. The price elasticity of the International Copper Cartel supply is then estimated using an econometric model, and found to be significantly larger in booms than in recessions. This supports the conclusion that members of the cartel cooperated during both booms and recessions. This finding raises the possibility that the International Copper Cartel of 1935–1939 actually could have enhanced social welfare by reducing price volatility, while also having little effect on the secular level of prices.

Appendix

Proof of proposition 1

By definition of price elasticity of demand Eq. (6) can be re-written as

$$ P\left({Q}^{nc},\theta \right)\left[1-\frac{Q_s^{nc}}{n{\varepsilon}_D{Q}^{nc}}\right]-{C_s}^{\hbox{'}}\left({Q}_s^{nc}/n\right)=0 $$

(15)

Taking the derivative of Eq. (15) with respect to price P, gives

$$ \left[1-\frac{Q_s^{nc}}{n{\varepsilon}_D{Q}^{nc}}\right]+{P}^2\frac{\partial }{\partial P}\left[1-\frac{Q_s^{nc}}{n{\varepsilon}_D{Q}^{nc}}\right]-\frac{{C_s}^{\hbox{'}\hbox{'}}}{n}\frac{\partial {Q}_s^{nc}}{\partial P}=0 $$

(16)

Multiplying the last expression by $P/{Q}_s^{nc}$, the price elasticity of supply for the strategic firms in the presence of the competitive fringe is obtained

$$ {\varepsilon}_S^s=\frac{{nC_s}^{\hbox{'}}}{Q_s^{nc}{C_s}^{\hbox{'}\hbox{'}}}+\frac{PK}{Q_s^{nc}{C_s}^{\hbox{'}\hbox{'}}{\varepsilon}_D}\left({\varepsilon}_D^s-{\varepsilon}_D\right), $$

(17)

where K is the strategic firms’ market share (i.e., ${Q}_s^{nc}/{Q}^{nc}$) and ${\varepsilon}_D^s$ is the price elasticity of demand faced by the strategic firms. Because in equilibrium, it can be shown that ${\varepsilon}_D^s=\frac{\varepsilon_D}{K}+\frac{\varepsilon_D}{K}{\varepsilon}_S^f$, the price elasticity of supply for strategic firms in the non-cooperative scenario is

$$ {\varepsilon}_S^{s, nc}=\frac{{nC_s}^{\hbox{'}}}{Q_s{C_s}^{\hbox{'}\hbox{'}}}+\frac{\varepsilon_S^{s, nc}}{n}\left(1+\frac{\varepsilon_S^f}{\varepsilon_D}\right)\frac{{C_f}^{\hbox{'}}}{{C_s}^{\hbox{'}}}\left(1-K\right) $$

(18)

Without a competitive fringe, it is not difficult to show that the former expression in the last formula is exactly the price elasticity of supply for the strategic group when no fringe is present, that is ${\varepsilon}_S^{s, nf}$, while the latter expression corresponds to the distortion in the price elasticity of supply introduced by the competitive fringe’s presence in the non-cooperative case, ${d}_f^{nc}$.

Finally, because in equilibrium $P\left[1-\frac{K}{\varepsilon_Dn}\right]-{C_s}^{\hbox{'}}=0$ and P = C _f ^' an alternative expression for the distortion introduced by fringe is

$$ {d}_f^{nc}=\frac{\varepsilon_S^{s, nf}}{n}\left(1+\frac{\varepsilon_S^f}{\varepsilon_D}\right)\left(\frac{{C_f}^{\hbox{'}}}{{C_s}^{\hbox{'}}}\left(1-n{\varepsilon}_D\right)+n{\varepsilon}_D\right) $$

(19)

Proof of proposition 2

In the cooperative case Eq. (10) can be re-written as

$$ P\left[1-\frac{{C_f}^{\hbox{'}\hbox{'}}{Q}_s^m}{{C_f}^{\hbox{'}\hbox{'}}{\varepsilon}_D{Q}^m+P}\right]-{C_s}^{\hbox{'}}=0 $$

(20)

Differentiation of (20) with respect to price P produces

$$ \left[1-\frac{{C_f}^{\hbox{'}\hbox{'}}{Q}_s^m}{{C_f}^{\hbox{'}\hbox{'}}{\varepsilon}_D{Q}^m+P}\right]+P\frac{\partial }{\partial P}\left[1-\frac{{C_f}^{\hbox{'}\hbox{'}}{Q}_s^m}{{C_f}^{\hbox{'}\hbox{'}}{\varepsilon}_D{Q}^m+P}\right]-\frac{{C_s}^{\hbox{'}\hbox{'}}}{n}\frac{\partial {Q}_s^m}{\partial P}=0 $$

(21)

Multiplying the last expression by $P/{Q}_s^m$, and using the fact that ${\varepsilon}_D^s$ is the price elasticity of demand faced by the strategic firms (see formula in Proof of Proposition 1), the price elasticity of supply for strategic firms in the cooperative scenario is

$$ {\varepsilon}_S^{s,m}=\frac{{nC_s}^{\hbox{'}}}{Q_s{C_s}^{\hbox{'}\hbox{'}}}+\frac{nP}{Q_s^m{C_s}^{\hbox{'}\hbox{'}}}\left[\frac{{C_f}^{\hbox{'}\hbox{'}}{Q}_f^m\left({\varepsilon}_D+{\varepsilon}_S^f\right)}{{C_f}^{\hbox{'}\hbox{'}}{\varepsilon}_D{Q}^m+P}+\frac{{C_f}^{\hbox{'}\hbox{'}}{C_f}^{\hbox{'}}\left({\varepsilon}_D+1\right)}{{\left({C_f}^{\hbox{'}\hbox{'}}{\varepsilon}_D{Q}^m+P\right)}^2}-\frac{{C_f}^{\hbox{'}\hbox{'}\hbox{'}}{\left({Q}_s^m\right)}^2{P}^2{\left({\varepsilon}_D\right)}^2}{{\left({C_f}^{\hbox{'}\hbox{'}}{\varepsilon}_D{Q}^m+P\right)}^3}\right], $$

(22)

The first term on the right-hand side of (20) is exactly the price elasticity of supply for the strategic group when no fringe is present, that is ${\varepsilon}_S^{s, nf}$, while the second term represents the distortion introduced in the cartel’s price elasticity of supply by the presence of non-cartel firms, ${d}_f^m$.