The AEO2006 projections of future energy market conditions reflect the effects of oil prices on the macroeconomic variables that affect oil demand, in particular, and energy demand in general. The variables include real GDP growth, inflation, employment, exports and imports, and interest rates. 

Although there is wide agreement that high oil prices have negative effects on U.S. macroeconomic variables, the magnitude and duration of the effects are uncertain. For example, most of the major economic downturns in the United States, Europe, and the Asia Pacific region since the 1970s have been preceded by sudden increases in crude oil prices. Although other factors were important, high oil prices played a critical role in substantially reducing economic growth in most of these cases. Recent history, however, tells a somewhat different story. Average world crude oil prices have increased by more than $30 per barrel since the end of 2001, yet U.S. economic activity has remained robust, growing by approximately 2.8 percent per year from 2001 through 2004. 

This section describes the ways in which oil prices affect the U.S. economy [17], presents a brief survey of the empirical literature on the economic impacts of changes in oil prices, and outlines the effects on the AEO2006 reference case projections of alternative assumptions in the high and low price cases. The results of the alternative cases indicate how the U.S. economy is likely to be affected by different levels of oil prices. 

Macroeconomic Impacts of High Oil Prices 

U.S. demand for crude oil arises from demand for the products that are made from it—especially gasoline, diesel fuel, heating oil, and jet fuel; and changes in crude oil prices are passed on to consumers in the prices of the final petroleum products. Increases in crude oil prices affect the U.S. economy in five ways: 

• When the prices of petroleum products increase, consumers use more of their income to pay for oil-derived products, and their spending on other goods and services declines. The extra amounts spent on those products go to foreign and domestic oil producers and, if wholesale margins increase, to refiners. Domestic producers may pay higher dividends and/or spend more on oil discovery, production, and distribution. Foreign producers may spend some or all of their extra revenues on U.S. goods and services, but the types of goods and services they buy will be different from those that domestic consumers would buy. How quickly and how much domestic and foreign oil producers spend on U.S. goods and services and financial and real assets will be critical in determining the effects of higher oil prices on the aggregate economy [18]. 
• Oil is also a vital input for the production of a wide range of goods and services, because it is used for transportation in businesses of all types. Higher oil prices thus increase the cost of inputs; and if the cost increases cannot be passed on to consumers, economic inputs such as labor and capital stock may be reallocated. Higher oil prices can cause worker layoffs and the idling of plants, reducing economic output in the short term. 
• Because the United States is a net importer of oil, higher oil prices affect the purchasing power of U.S. national income through their impact on the international terms of trade. The increased price of imported oil forces U.S. businesses to devote more of their production to exports, as opposed to satisfying domestic demand for goods and services, even if there is no change in the quantity of foreign oil consumed. 
• Changes in oil prices can also cause economic losses when macroeconomic frictions prevent rapid changes in nominal prices for final goods (due to the costs of changing “menu” prices) or for key inputs, such as wages. Because there is resistance on the part of workers to real declines in wages, oil price increases typically lead to upward pressure on nominal wage levels. Moreover, nominal price “stickiness” is asymmetric, in that firms, unions, and other organizations are much more reluctant to lower nominal prices and the wages they receive than they are to raise them. When a nominal increase in oil prices threatens purchasing power, the adjustment process is slowed, with multiplier effects throughout the economy [19]. 
• Finally, higher oil prices cause, to varying degrees, increases in other energy prices. Depending on the ability to substitute other energy sources for petroleum, the price increases can be large and can cause macroeconomic effects similar to the effects of oil price increases. 

The nature of the oil price increases, the state of the economy, and the macroeconomic policies undertaken at the time may accentuate or dampen the severity of adverse macroeconomic effects. If price increases are large and sudden, their impacts on short-term growth may be much larger than if they are gradual, because sudden oil price shocks scare households and firms and prevent them from making optimal decisions in the near term. 

On the potential output side, sudden large price increases create widespread uncertainty about appropriate production techniques, purchases of new equipment and consumer durable goods like automobiles, and wage and price negotiations. As firms and households adjust to the new conditions, some plant and equipment will remain idle, some workers will be temporarily unemployed, and the economy may no longer operate along its long-run production-possibility frontier. Although it is easy to differentiate gradual from rapid price increases on a conceptual basis, empirical differentiation is more difficult. 

In terms of the state of the economy, if the economy is already suffering from high inflation and unemployment, as in the late 1970s, then the oil price increases have the potential to cause severe damage by limiting economic policy options. Many analysts assert that it was the monetary policy undertaken in the 1970s that really damaged the U.S. economy. 

The economic policies that are followed in response to a combination of higher inflation, higher unemployment, lower exchange rates, and lower real output also affect the overall economic impact of higher oil prices over the longer term. Sound economic policies may not completely eliminate the adverse impacts of high oil prices described above, but they can moderate them. Conversely, inappropriate economic policies can exacerbate the adverse impacts. Overly contractionary monetary and fiscal policies to contain inflationary pressures can worsen the recessionary effects on income and unemployment; expansionary monetary and fiscal policies may simply delay the fall in real income necessitated by the increase in oil prices, stoke inflationary pressures, and worsen the impact of higher prices in the long run. 

Empirical Studies of Oil Price Effects 

The mechanism by which oil prices affect economic performance is generally well understood, but the precise dynamics and magnitude of the effects are uncertain. Quantitative estimates of the overall macroeconomic damage caused by oil price shocks in the past and of the economic gains realized by oil-importing countries as a result of the oil price collapse in 1986 vary substantially, in part because of differences in the models used to examine the issue [20]. Two different approaches have been used to estimate the magnitude of oil price effects on the U.S. economy. One uses large, disaggregated macroeconomic models of the economy, and the other uses time-series analysis of historical events to estimate directly the macroeconomic effects of oil price changes. 

In the first approach, macroeconomic models are used in attempts to account for all the relationships among the major macroeconomic variables in the economy (as described by the National Income and Product, Balance of Payments, and Flow of Funds Accounts), and historical data are used to estimate statistically the parameters linking the variables. The advantages of macroeconomic models are consistent accounting of macroeconomic relationships over time and the ability to account for other events taking place. 

A recent Stanford University Energy Modeling Forum (EMF) study by Hillard Huntington found that most macroeconomic models report similar economic effects of oil price increases [21]. Table 9 shows the results for real GDP, the GDP price deflator, and unemployment obtained from three models and their averages [22]. The results are shown for a 33-percent increase in the oil price, from $30 to $40. For example, the output results in Table 9 imply that a 33-percent increase in the oil price sustained for 2 years reduces real GDP relative to the baseline by 0.2 percent in the first year and 0.5 percent in the second year. In terms of an elasticity response of real GDP to oil price, the percentage change in real GDP relative to the percentage change in oil price is approximately 0.01 in the first year and 0.02 in the second year. 

The second approach is simpler, focusing specifically on the relationship between changes in crude oil prices and some measure of their economic impact, such as aggregate output, inflation, or unemployment. Time-series analyses of historical data are used to estimate statistically an equation (or a system of equations called “vector autoregressions”) that explains economic growth rates as a function of the past growth in the economy and past changes in crude oil prices. Many studies add the past values of additional variables to the system in order to incorporate their interactions with the oil price and GDP variables. 

Table 10 shows results for the U.S. economy from a recent study by Jimenez-Rodriguez and Sanchez [23], which are representative of the results obtained in the time-series literature. Due to the nature of the reduced-form framework used, the results are direct estimates of GDP elasticities with respect to oil price changes as of the given quarter after the permanent price change. The asymmetric results allow separate estimates of GDP elasticity for oil price increases, decreases, and net increases (when oil prices exceed the maximum over the previous 12 quarters). When the six-quarter GDP elasticity estimated by Jimenez-Rodriguez and Sanchez (approximately 0.05) is applied to a 33-percent price increase (to be comparable with the average macroeconomic simulation response in Table 10), real GDP declines by 1.7 percent—more than 3 times the effect on real GDP in macroeconomic simulations. 

figure data

Generally, as indicated by the results in Table 10, time-series studies show larger impacts on output and other variables than do macroeconomic simulations. Huntington offers four major reasons as to why the empirical estimates are so different: 

• The larger impacts calculated from direct statistical estimations often are attributed to a range of macroeconomic frictions that could make the economy’s response to an oil price shock fundamentally different from its response to a smaller increase in oil prices. Large macroeconomic models do not differentiate between oil price increases and decreases, or between surprise events and more gradual price adjustments. 
• The larger estimates from time-series models may also reflect baseline economic conditions before an oil price disruption that are fundamentally different from today’s economic environment. For example, the oil price shocks of the 1970s hit the U.S. economy when it already was experiencing inflationary pressures. 
• Historical oil price shocks reduced not only aggregate output but also the country’s purchasing power. Real national income fell as the costs of buying international goods (including oil) increased more than income from exports. The higher prices made the country poorer by requiring more exports to balance each barrel of imported oil, leaving less aggregate output for domestic consumption. 
• The oil price shocks of the 1970s completely surprised firms and households in many different countries at the same time. Firms and households made decisions about production and prices that had important consequences for the strategies of other firms in the economy [24]. And yet, there was little opportunity to coordinate strategies in such an uncertain world. Now, after several different oil price episodes, there has been significant learning about how to cope with the uncertainties created by oil price shocks. It is unlikely that firms and households will be surprised in the same way or to the same degree as they were by earlier shocks. 

If crude oil prices rise early in a particular year, what will be the impact on the economy at the end of the following year? Huntington offers the following tentative answers, and Table 11 summarizes the impacts on GDP, as well as the impacts on the GDP price deflator for all goods and services and the unemployment rate. If the economy is operating at its potential output level and inflation is constant, a reasonable estimate is that a 10-percent increase in the price of oil that does not surprise households and firms (higher oil price in Table 11) will reduce potential output (GDP) by 0.2 percent. If the economy is operating well below its potential output level, the impact on GDP may be somewhat larger but is unlikely to exceed 0.2 percent after the first year. If the oil price increase comes as a complete surprise and the economy is already in a rising inflationary environment (oil price shock in Table 11), then it has the potential to cause larger economic losses, which would be closer to those predicted by time-series models. 

AEO2006 Price Cases 

The key feature of the AEO2006 high and low world oil price paths is that they are not characterized by disruption, but rather represent a gradual and sustained movement relative to the reference case path. Keeping this distinction in mind, the Macroeconomic Activity Module in NEMS, which contains the Global Insight Inc. (GII) Macroeconomic Model, is used to assess the economic impacts of the alternative price paths. 

Most of the results projected for the U.S. economy in the high and low price cases relative to the reference case are similar to the results for macroeconomic models discussed above. The AEO2006 high and low price cases are unique, however, in that they trace out, in a consistent manner, both the short-term impacts of oil price increases and the longer term adjustments of the economy in response to sustained high and low prices by employing a disaggregated macroeconomic model integrated with a very detailed energy market model—NEMS. 

Figure 12 shows the percentage change from the reference case projections for real GDP and oil prices in the AEO2006 high and low price cases. In the high price case, oil prices rise rapidly to 70 percent above reference case prices within 10 years (2016), then climb more gradually to 80 percent above reference case prices in 2030. In the low price case, oil prices do not change by as much relative to the reference case, declining to 34 percent below reference case prices in 2016 and 44 percent below in 2030. Consequently, the macroeconomic effects in the two cases are not expected to be symmetric. 

In each of the three cases, the U.S. economy grows at an average annual rate of 3.0 percent from 2004 through 2030 (although the average growth rates in the three cases do differ when calculated to two or more decimal places). With such significant differences in oil price paths in the three cases, why is the impact on the long-term real GDP growth rate so small? The major reasons have to do with the nature of the oil price increases and decreases relative to the reference case and their short-term versus long-term impacts on the economy. 

The oil price projections for 2005 and 2006 are the same in the three cases. From 2007 to 2010, the real oil price increases by more than 2 percent annually in the high price case, declines by 5 percent annually in the reference case, and declines by 9.4 percent annually in the low price case. From 2010 to 2015, the annual changes in oil prices in the three cases average 4 percent, -0.5 percent, and -5 percent, respectively. After 2015 the differences narrow considerably, and by 2030 the annual increases in oil prices average 1.1 percent in the high price case, 0.8 percent in the reference case, and zero in the low price case. With the maximum differences in growth rates among the three cases occurring in 2010, the peak impacts on real GDP and other economic variables occur approximately 2 years later, in 2012. 

Over the 2006-2030 period, real GDP in the high price and low price cases deviates from that in the reference case for a considerable period. As the economy adjusts to the oil price changes, however, the differences become smaller, and by 2030 real GDP is approximately the same in the three cases, at $23,112 billion in the reference case, $23,054 billion in the high price case, and $23,178 billion in the low price case. 

The discounted sum of changes in real GDP over the entire projection period provides a better indicator of net effects on the economy. In the low price case, the sum of the changes in real GDP, discounted at a 7-percent annual rate, over the 2006-2030 period is $665 billion, and in the high price case the sum is -$869 billion. These sums represent approximately 0.4 percent and -0.5 percent, respectively, of the total discounted real GDP in the reference case over the same period. 

figure data

The elasticity of real GDP with respect to oil price changes over the 2006-2030 period is -0.007 in both the high price and low price cases. The year-by-year (marginal) and up-to-the-year (average) elasticities of real GDP with respect to oil price changes in the high price case (Figure 13) shows that the short-term effects of oil price increases are larger than their long-term effects. 

To portray the short-term dynamics of the economy as it reacts to oil price changes, Table 12 shows 5-year average annual growth rates for U.S. oil prices (the imported refiners acquisition cost of crude oil), real GDP, potential GDP, and the consumer price index (CPI), as well as 5-year averages for the Federal funds rate and unemployment rate, over the 2005-2030 period. Higher oil prices in the short term feed through the economy and reduce aggregate expenditures on goods and services. As aggregate demand is less than aggregate supply, unemployment increases. 

With higher prices there would also be a tendency for interest rates to rise. In the high price case, real GDP growth averages 3 percent per year over the 2005-2010 period, CPI inflation averages 2.3 percent per year, and the average unemployment rate for the 5-year period is 5 percent. In the reference case, the comparable rates are 3.2 percent (average annual real GDP growth), 2 percent (average annual CPI inflation), and 4.8 percent (unemployment). Potential GDP growth and the Federal funds rate are not significantly different in the two cases over the 2005-2010 period. The impacts of high prices on real GDP shown in Table 12 are in agreement with the average results shown in Table 9. 

In the high price case, as unemployment increases, the Federal Reserve lowers the Federal funds rate from its projected level in the reference case. At the same time, total employment costs are lower, which tends to slow price growth in the economy. Over the 2010-2015 period, even though oil prices continue to grow by 4.1 percent annually in the high price case (as opposed to declining by 0.5 percent annually in the reference case), real GDP growth is about the same in the two cases, although it is increasing from a lower base in the high price case. The Federal funds rate is lower in the high price case than in the reference case, and the unemployment and CPI inflation rates are higher. 

After 2015, as the differential in the oil price growth rates between the high price and reference cases shrinks, rebound effects from the lower employment costs and lower Federal funds rate in the high price case are stronger than the contractionary impacts of higher oil prices, leading to higher real GDP growth and lower CPI inflation than in the reference case. As a result, in 2030, the real GDP growth rate and unemployment rate in the high price case are nearly the same as in the reference case, but the Federal funds rate is lower. 

The assumptions behind the oil price cases are that: the price changes do not come as a shock and come to be expected over time; the Federal Reserve is able to carry out an activist monetary policy effectively, because core inflation remains low; exchange rates do not change from those in the reference case; and other countries experience impacts similar to those in the United States. Changes in any of these assumptions could increase the projected impacts on the U.S. economy. 

The economic impact of oil price changes is an issue that continues to attract considerable attention, especially at this time, when oil prices have continued to rise over the past 3 years. Over the past 30 years, much has been learned about the nature of the economic impacts and the extent of damage possible. Empirical estimates based on history provide two sets of results. In the 1970s and 1980s the damages were substantial, and it is believed that recession followed—and may have been caused by—the oil price increases. Current literature suggests that, in today’s U.S. economy, sustained higher oil prices can slow short-term growth but are not likely to cause a recession unless other factors are present that shock economic decisionmakers or lead to inappropriate economic policies. The AEO2006 high and low price cases provide estimates of the economic impacts on such an economy, and the projections in the price cases are within the range that other macroeconomic models predict.

Notes and Sources

Contact: Nassir Khilji
Phone: 202-586-1294
E-mail: nassir.khilji@eia.doe.gov