Since the start of the Covid pandemic in 2020 (IMF 2024), fluctuations in oil and other commodity prices have played a key role in the surge and decline in inflation. The impact of supply bottlenecks on post-Covid goods prices, and the impact of Russia’s invasion of Ukraine on energy prices, has raised interest in the relationship between consumer prices and commodity prices. Some recent contributions include Killian and Zhou (2023) and Coutinho and Licchetta (2024). However, this is not a new phenomenon. The importance of dramatic fluctuations in crude oil prices due to increased consumer prices has been debated since the oil crisis of the 1970s.
A recent paper (Gerlach and Stuart 2024) looks back further and shows that the commodity price shock had already had a major impact on widespread economic inflation before World War I. Importantly, it shows that the swings in the global business cycle caused fluctuations in product demand, which had a particularly strong impact on consumer price inflation during this period.
To study this question, annual data on British commodity prices and local currency consumer prices for 15 economies for the period 1851-1913. Consumer price data can be seen as obtaining costs of living from a variety of sources. Our product prices are sourced from Saalbeck (1886, 1893, 1908, and 1917), where data on the price of 43 products is compiled. One of the problems with the past consumer price series is that wholesale prices are often used as a proxy for retail prices. Therefore, we do not limit our analysis to metals or products that do not fit into the consumer basket, nor use it as a proxy for items within the consumer basket.
Figure 1 shows the median international inflation rates for 15 economies, along with median commodity price inflation. The product price is more unstable than the consumer price, but the two series move together.
Figure 1 Median product price (11 products) and median inflation rate (15 countries), annual change, 1851-1913
First, we use regular least squares (OLS) to estimate the relationship between delayed inflation and national inflation regarding median growth rates for a subset of the above products. The product price measure is important at the 5% level in all cases except Australia (see Table 5 in the paper for a complete set of results). This result is surprising when we recall that there are no direct channels that could affect consumer prices, as these are unlikely to be included in the consumer basket. The estimated coefficient ranges from 0.09 in France to 0.60 in Switzerland, with an average of 0.26, suggesting that more than a quarter of the change in commodity price will transition to inflation within a year. The percentage of inflation explained by the model ranges from 8% in Australia to 48% in the US, with an average of 24%.
An important theme in the literature is that commodity prices strongly move with the global business cycle and demand factors (Pindyck and Rotemberg 1990, Alquist etal. 2020, Delle etal. 2022). One important issue is whether the strength of the commodity price shock transmission reflects the global boom and demand for the product, or whether it reflects product-specific supply factors. To investigate this issue, we will distinguish between changes in commodity prices induced by supply and demand.
Figure 2 shows Crafts et al. (1989). The correlation between these two series is 0.43. The shaded area in Figure 2 shows the period of the UK recession (based on Klovland 1998). The relationship is not perfect, but commodity prices tend to fall during slump and rise during expansion.
Figure 2 Median change in product prices (11 products), and the UK Industrial Production and Business Cycle Chronology, 1851-1913
The UK was one of the most important economies of the time, but in our case, a broader measure of the international aspect of the business cycle is desirable. Klovland (2003) shows that shipping freight charges earn movement in the global business cycle. Measure freight rates using median fees from several sources (Klovland 2003, North 1958, Harley 1988) and deflate them using the UK GDP deflator. Figure 3 shows that there was a strong symbiosis between the state of the global business cycle and this measure of commodity price.
Figure 3 Median product price and freight rate, annual change, 1851-1913
Commodity prices to assess whether the supply and demand movement of commodity prices have the same effect on inflation. If demand-induced increases in commodity prices have a greater impact on inflation than changes in price due to supply, then the IV estimate of γI is expected to be greater than the OLS estimate.
If the equation is estimated by OLS, we can see that the average parameter of the commodity variable is 0.35 when IV is used in contrast to 0.26 (the complete set of IV parameters is used in Table 6 of the paper can). The estimated parameters using IV range from 0.01 in the US to 0.84 in Switzerland. Efficiency is always an issue when using IV. Nevertheless, this parameter is important at the 5% level in 9 out of 15 cases and at the 10% level in 2 more. These results are compatible with the idea that demand-driven fluctuations in commodity prices are more strongly reflected in the economic inflation we are studying.
We use the Hausman test of endogeneic bias to show that the IV estimate is generally greater than the OLS estimate.
Overall, commodity price shocks can be seen as a key determinant of the international joint movement of inflation before World War I, particularly when it was due to changes in international demand.
reference
Alquist, R, S Bhattarai and O Coibion (2020), “Commonity Price Collaboration and Global Real Activities”, Journal of Monetary Economics 112:41-56.
Coutinho, L and M Licchetta (2024), “Inflation differentials of the euro region at the time of higher energy prices,” Voxeu.org, February 22nd.
Crafts, NFR, SJ Leybourne and TC Mills (1989), “Trends and Cycles of British Industrial Production, 1700-1913”, Journal of the Royal Statistical Society, Series A, 152(1): 43-60.
Delle Chiaie, S, L Ferrara and D Giannone (2022), “Common Factors of Commodity Prices”, Journal of Applied Econometrics 37(3): 461-476.
Gerlach, S and R Stuart (2024), “Product Prices and International Inflation, 1851-1913”, Journal of International Money and Finance 144, and CEPR Discussion Paper DP16526.
Harley, CK (1988), “Marine Cargo Cargo Rate and Productivity, 1740-1913: Reaffirming the Advantages of Mechanical Invention”, Journal of Economic History 48:851-876.
IMF (2024), “The Great Tightening: Insights from Recent Inflation Episodes,” Chapter 2 World Economic Outlook, October, 41-64.
Kilian, L and X Zhou (2023), “The impact of inflation on energy prices,” Voxeu.org, February 10th.
Klovland, J.T. (1998), “Reevaluation of the UK Business Cycle Chronology 1850–1914”, Tjo Dick (ed.), Business Cycles since 1820: A New International Perspective from Historical Evidence, 49–90, Edward Elgar.
Klovland, J.T. (2003), “Business Cycle, Commodity Prices, and Shipping Freight Fees: Some Evidence from Pre-WWI Periods”, SNF Report No. 48/02.
North, D (1958), “Marine freight rates and economic development, 1750-1913”, Journal of Economic History 18:537-555.
Pindyck, R and J Rotemberg (1990), “Over-cooperative movement of commodity prices”, Economic Journal 100:1173–89.
Sauerbeck, A (1886), “Prices for Goods and Precious Metals”, Journal of the Statistical Society of London 49(3): 581-648.
Sauerbeck, A (1893), “Product Prices over the past seven years”, Journal of the Statistical Society of London 56(2): 215-254.
Sauerbeck, A. (1908), “Product Prices for 1907”, Journal of the Statistical Society of London 71(1): 132-146.
Sauerbeck, A. (1917), “Wholesale Prices for Products 1916”, Journal of the Statistical Society of London 80(2): 289-309.
Staiger, D and J.H. Stock (1997), “Regression of instrumental variables with weak instruments,” Econometrica 65(3):557-586.
