ESTUDIOS SOBRE LA ECONOMÍA ESPAÑOLA

Oil Shocks and the Business Cycle in Europe Carlos De Miguel Baltasar Manzano José Mª Martín-Moreno

EEE 215

October 2005

ISSN 1696-6384

Las opiniones contenidas en los Documentos de la Serie EEE, reflejan exclusivamente las de los autores y no necesariamente las de FEDEA.

The opinions in the EEE Series are the responsibility of the authors an therefore, do not necessarily coincide with those of the FEDEA.

Carlos De Miguel, Baltasar Manzano and José Mª Martín-Moreno

Universidad de Vigo and rede

Abstract

This paper analyzes the effects of oil price shocks on the business cycle of the EU-15 countries using a standard dynamic general equilibrium model for a small open economy in which oil is included as an imported productive input. The results show that oil shocks can account for a significant percentage of GDP fluctuations in many of those countries. Furthermore, we show that the increases in the relative price of oil had a negative effect on welfare, particularly in southern European countries, which are historically associated with a lax monetary policy during oil crisis.

Key words: Oil crisis, aggregate fluctuations, welfare cost. JEL: Q43, E32.

^*The authors would like to thank participants in the 1^st Atlantic Workshop on Energy and Environmental Economics for their valuable suggestions. Financial support from the Spanish Ministry of Science and Technology and from FEDER through grant BEC2002-01995 and from Xunta de Galicia (PGIDIT03PXIC30001PN, PGIDIT03CSO30001PR) is gratefully acknowledged.

cmiguel@uvigo.es; bmanzano@uvigo.es; jmartin@uvigo.es

Universidad de Vigo. Lagoas-Marcosende s/n. 36310 Vigo (Spain).

1.- Introduction

The effects of oil price changes on economic activity have been widely studied. An increase of oil prices tends to reduce the level of economic activity, given its implications on the evolution of important macroeconomic variables and due to the strong dependence of western economies on this input.

Recently, we have witnessed a substantial increase in the prices of oil, derived mainly from the high demand of emerging economies (China among others) and from the political and economic situation of Iraq. In this framework, the economic problems derived from trends in oil prices appear again, not only with regard to the situation of the markets, but also to their possible macroeconomic effects. Tensions in the oil markets have returned, and we are perhaps entering a period that some experts call the end of the cheap oil age. The ghost of old crises seems to be reappearing, stirring a fear of recessions in the main economies that integrate the European Union.

As it is very well-known, for economies so dependent on oil, increases in the price of the barrel have different adverse effects. In the short run, they involve a reduction of GDP and an increase of the inflation rate. In the medium and long run, industrial production is affected, consumption decreases due to the fall of purchasing power and investment also falls, affecting the cyclical position of the economy and citizens´ welfare. Consequently, with oil prices between 40 and 50 dollars per barrel, the economic implications should be studied.

However, to be able to carry out a rigorous analysis of what this new oil crisis in the European setting can represent, and to make an accurate analysis of economic effects, we must situate ourselves in a certain time frame. Thus, in this paper, we analyze the effects of the increase in oil prices since the first crisis (70s), on economies so dependent on oil as the Europeans’.

The goal of the article is twofold. Firstly, we analyze the importance of this source of shocks for both the size and the shape of aggregate fluctuations in those economies during the last three decades. Secondly, we quantify the effects of the changes in relative oil prices on welfare in these economies.

The effects of oil price shocks on industrialized economies have been widely acknowledged in economic literature. Pindyck (1979), Hamilton (1983) and Olson (1988) suggest that these shocks affect growth as well as the business cycle, thereby becoming an additional source of economic fluctuation. There is extensive empirical literature that offers evidence of an asymmetric relationship between oil prices and aggregate economic activity (see Mork, 1989 and 1994). Kim and Loungani (1992) and Finn (1995) analyze the role of energy price shocks using real business cycle models in closed economies, focusing on the US. These authors find that such shocks offer very little help in explaining the aggregate fluctuations in the economy in question. However, De Miguel, Manzano and Martín-Moreno (2003) show that, in a small open economy framework, oil price shocks are very important when explaining aggregate fluctuations.

This paper presents a real business cycle model for the different European economies. The model used is a standard general equilibrium model of a small open economy in which oil is included as an imported productive input. The relative oil price as well as the real interest rate is assumed to be set in international markets, so we consider a small open economy in the sense of taking those prices as given. Oil price shocks are the only source of fluctuation considered. Therefore, although the economy is hit by many shocks, our analysis is conditional on a single shock. Thus, this analysis would allow us to verify the extent to which oil price shocks can account for aggregate fluctuations in European economies.

The results show that oil shocks can account for a significant percentage of GDP fluctuations in many of those countries, although the explanatory power is smaller for others. This wide range of variation can be explained by differences in the strength of monetary policies that affect relative oil prices in each country. In addition, the model reproduces the cyclical path of the European economies in periods of oil crisis. Finally, we show that increases in the relative price of oil had a negative effect on welfare, particularly in southern European countries, which are historically associated with a lax monetary policy during oil crises.

The remainder of this paper is organized as follows. Section 2 describes the model. Section 3 discusses the choice of parameter values. We report the main results in Sections 4 and 5. Finally, in the last section, we present the conclusions.

2.- The Model

The model described in this section is a stochastic dynamic general equilibrium model of a small open economy populated by a large number of infinite-lived households, and firms that need to import oil to produce a consumption good. The basic structure of the model, in terms of preferences and technology, is similar to the De Miguel, Manzano and Martín-Moreno (2003) structure. The general features of the model are the following:

The production of the final good, requires the use of labor, capital, and energy, The production technology of firms is described by a nested CES function with constant returns to scale:

\[F (n _ {t}, k _ {t}, e _ {t}) = n _ {t} ^ {\theta} \left[ (1 - a) k _ {t} ^ {- \upsilon} + a e _ {t} ^ {- \upsilon} \right] ^ {- \frac {1 - \theta}{\upsilon}},\tag{1}\]

where is the labor share and the parameter υ is equal to , where s is the elasticity of substitution between capital and energy.

The economy’s resource constraint for period t is given by:

\[c _ {t} + i _ {t} + x n _ {t} = y _ {t},\tag{2}\]

where is private consumption, is investment and are net exports. Capital, accumulates according to the law of motion:

\[i _ {t} = k _ {t + 1} - (1 - \delta) k _ {t} + \Phi (k _ {t}, k _ {t + 1}),\tag{3}\]

where is the depreciation rate and Φ is a capital adjustment cost function which we assume to be quadratic:

\[\Phi (k _ {t}, k _ {t + 1}) = \frac {\phi}{2} \left(\frac {k _ {t + 1} - k _ {t}}{k _ {t}}\right) ^ {2}.\tag{4}\]

The representative firm solves:

\[M a x F (n _ {t}, k _ {t}, e _ {t}) - w _ {t} n _ {t} - r _ {t} k _ {t} - p _ {t} e _ {t}, \forall t\tag{5}\]

where is the wage, the capital rate of return and the relative oil price. In equilibrium, marginal productivities are equal to input prices: and

The relative oil price follows a stationary stochastic process:

\[\ln p _ {t} = \overline {{p}} + \rho \ln p _ {t - 1} + \varepsilon_ {t}, \qquad \varepsilon_ {t} \sim N (0, \sigma_ {p}), \quad | \rho | < 1.\tag{6}\]

We assume that the individuals can buy or sell an international asset, at an exogenous international interest rate, . The evolution of net exports is dictated by:

\[x n _ {t} = p _ {t} e _ {t} + b _ {t + 1} - (1 + r ^ {*}) b _ {t,},\tag{7}\]

where are oil purchases.

Consumers maximize the expected value of lifetime utility subject to their budget constraint:

\[\begin{array}{l} \text {Max} E _ {0} \Bigg \{\sum_ {t = 0} ^ {\infty} \beta^ {t} \frac {1}{1 - \sigma} \bigg [ (c _ {t} - \psi n _ {t} ^ {\nu}) ^ {1 - \sigma} - 1 \bigg ] \Bigg \} \\ \text {s.t.} c _ {t} + k _ {t + 1} - (1 - \delta) k _ {t} + \Phi (k _ {t}, k _ {t + 1}) + b _ {t + 1} = w _ {t} n _ {t} + r _ {t} k _ {t} + (1 + r ^ {*}) b _ {t}, \end{array}\tag{8}\]

where is the subjective rate of intertemporal discount, is the parameter of relative risk aversion, ν is one plus the inverse of the intertemporal elasticity of substitution of labor supply and _ψ is a positive parameter.

The conditions that solve the consumer's problem are the following:

\[U _ {n _ {t}} + U _ {c _ {t}} w _ {t} = 0,\tag{9}\]

\[U _ {c _ {t}} \left(1 + \phi \frac {k _ {t + 1} - k _ {t}}{k _ {t} ^ {2}}\right) = \beta E _ {t} \left\{U _ {c _ {t + 1}} \left(1 - \delta + r _ {t + 1} + \phi \frac {k _ {t + 2} - k _ {t + 1}}{k _ {t + 1}} \frac {k _ {t + 2}}{k _ {t + 1} ^ {2}}\right) \right\},\tag{10}\]

\[U _ {c _ {t}} = \beta E _ {t} \left\{U _ {c _ {t + 1}} (1 + r ^ {*}) \right\},\tag{11}\]

\[c _ {t} + k _ {t + 1} - (1 - \delta) k _ {t} + \Phi (k _ {t}, k _ {t + 1}) + b _ {t + 1} = w _ {t} n _ {t} + r _ {t} k _ {t} + (1 + r ^ {*}) b _ {t}.\tag{12}\]

3.- Parameter values

We now briefly describe our procedures for selecting parameter values listed in Table 1. We follow the standard real business cycle literature in using steady-state conditions to find parameter values matching average values observed in the data, while other parameters will be equal to standard values used in the literature. The model is calibrated to reproduce average values of the European Union in annual data from 1960- 2003, before the unification of 2004 (EU-15). The main sources for the data used are the AMECO Database from Eurostat and International Energy Agency (IEA) Statistics.

The depreciation rate of capital is obtained from equation (3) in steady state, , where i/k is the average value of EU-15 data, while the discount factor was set by using equation (11) in steady state, , so as to give a real annual interest rate of 4%.

Table 1.- Parameters of the economy.

The value of representing the importance of oil with respect to capital in the production function, is obtained from the first order conditions of the firm’s problem in the steady state, , where and represent average values over the sample, and υ is borrowed from Kim and Loungani (1992).

Table 2.- Oil Price Process.

The value of the parameter is chosen from (9) in the steady state: , assuming that the productive time is 5476 hours per year. The output per worker, and the labor share, represent averages for the EU-15 economies.

The parameter of the adjustment costs is calibrated so the variability of investment relative to output in data would be reproduced by the model.

The remaining parameters are chosen in conformity with earlier studies. The parameter of the utility function v is taken from Greenwood et al. (1988), while the risk aversion parameter σ is obtained from Mendoza (1991).

Finally, the stochastic process parameters of relative oil prices for each country of the EU-15 are estimated, in domestic currency, from equation (6). Table 2 reports the results of the estimation.

4.- Oil shocks and EU-15 aggregate fluctuations

In this section, we test how accurately the model driven by oil price shocks can fit the business cycle of the EU-15 between 1970 and 2003. We start by running simulations, including for each country the corresponding stochastic process for the relative oil price presented in table 2. Such an experiment allows us to compare the actual GDP data with the corresponding fluctuations of the output in the model, obtaining the percentage of GDP volatility that can be explained by the model with oil shocks. Thus, we can explore to what extent output fluctuations at each European country could be generated by oil shocks. The simulation results are summarized in table 3.

The relative price of oil in each country is obtained as the ratio between the price of the Brent barrel expressed in domestic currency and the corresponding GDP deflator. Therefore, the differences of relative oil prices between countries arise either from inflation or from the exchange rate. In this sense, both the exchange rate and the monetary policies would have been the main tool to accommodate the effects of oil crises in each country.

In light of the simulation results, we can consider several groups of countries. The first group includes Portugal, Spain, Greece and Italy, that is, countries with large GDP fluctuations, where the role of oil shocks is remarkable, responsible from 30% to 42% of total output fluctuations in Portugal and Italy, respectively. Those results are consistent with the common view about southern European countries’ excessively lax monetary policy, which led to more difficulties in accommodating oil shocks.

There is another group of countries (Austria, Denmark and Sweden) in which oil disturbances also play an important role, explaining around 30% of output fluctuations, although their output is quite less volatile.

Table 3.- Comparison between the Predictions of the Model and the Data.

The main group includes the rest of the EU-15 countries, in which the contribution of oil shocks to explain GDP fluctuations scores from 10% to 20%. That group includes countries like France, UK, Netherlands, Belgium, Luxembourg and Germany, whose central banks have implemented stronger monetary policies to face the oil crises of the 70s.

Therefore, as it was pointed out above, the role of monetary policies has been crucial in accommodating the oil shocks that hit western economies in the last three decades.

Oil shocks are an important source for explaining aggregate fluctuations across Europe, although with a wide range of variation. But it is also important to test whether the introduction of oil shocks can mimic the business cycle shape. In order to do that, we simulate the model with the actual path of the relative price of oil as the only source of fluctuation. The analysis allows us to verify the extent to which oil prices actually give rise to the business cycle path of the EU-15 countries.

Figures 1-15 represent the evolution of the relative price of imported oil, and the comparison between the evolution of the actual GDP and the output path simulated by the model for each country. Given that we focus on the business cycle, both series have been detrended using the Hodrick-Prescott filter.

Figure 1. Portugal.

Figure 2. Spain.

Figure 3. Greece. Figure 4. Italy.

Figure 5. France.

Figure 6.United Kingdom.

Figure 7. Ireland. Figure 8. Germany.

Figure 9. Belgium.

Figure 10. Luxembourg. Figure 11. Netherlands.

Figure 12. Austria.

Figure 13. Sweden.

Figure 14. Finland.

Figure 15. Denmark.

The chart of the relative oil prices roughly shows the same path for all countries analyzed. The first great oil crisis was triggered by the Arab-Israeli Yom Kippur War when the OPEC imposed an oil embargo on western countries, thereby dramatically increasing oil prices at the beginning of 1974. The increase had immediate effects on the evolution of the output in the model as well as on the GDP data, though with some delay in several countries. Therefore, the model satisfactorily reproduces the negative effect on output during the first oil crisis. The next period which calls for our attention is the second oil crisis, which takes us back to 1979 and was triggered by the Iranian

Revolution, causing the relative price of oil to double in less than a year. Again, the increase in oil prices reduces the output in both the model and the economy in most countries, which is accurately depicted by the model. Some countries responded to the oil crisis with some delay (Germany, Belgium, Netherlands, Austria and Italy). At the beginning of 1986, the oil market collapsed, bringing about a significant drop in oil prices. This situation affected output in a positive way, which quickly recovered. The model reacts instantly when confronted with the fall in oil prices, while the recovery in the data begins some periods later in most countries.

Summing up, the model guided by changes in oil prices is able to mimic the shape of the output fluctuations in episodes of dramatic changes in the oil market, especially for the two crises in the 70s and the situation in the mid-80s. However, when oil market conditions were stable, as they became in 1986, there are other disturbances that explain aggregate fluctuations, so the model driven by oil shocks fails to reproduce the cyclical path of the countries analyzed.

The results show the vulnerability of small open economies heavily dependent on imported oil when they are confronted with large changes in the conditions that control the international oil market. This highlights the importance of considering the behavior of relative oil prices when analyzing the business cycle.

5.- Oil Price Shocks and Welfare

The previous section stresses the influence of oil shocks on both the size and the shape of aggregate fluctuations. Nevertheless, oil shocks affect not only GDP fluctuations but also welfare. As De Miguel, Manzano and Martín-Moreno (2003) emphasized, those effects are particularly important in the framework of a small open economy dependent on oil imports because of no domestic oil production, and the possibilities of substituting this input with capital are limited due to the complementarity between capital and oil. In this section, we evaluate the loss of welfare occasioned by oil price increases registered from the first half of the 70s until the second half of the 80s. The relative price in the last quarter of 1973 was very similar to the value registered in the third quarter of 1986 in each country; therefore, we can consider the intermediate period as a temporary price increase whose welfare cost we calculate.

The welfare loss of oil crises is estimated by the welfare cost derived from the increase in oil prices with respect to the initial situation. This cost is defined at each moment as the percentage of the increase in consumption that an individual would require to enjoy the same level of welfare with respect to the starting point. The welfare cost at each moment is calculated through the x variable that solves the following equation:

\[\overline {{U}} = U \bigl [ \widetilde {c} (1 + x), \widetilde {n} \bigr ],\tag{13}\]

where is the level of utility reached at the initial situation, in this case 1973, while and represent consumption and hours worked at each moment, which provide a level of welfare U. Thus, the product indicates the total increase in consumption required to restore the initial level of welfare. This welfare cost measure is usually expressed as a percentage of output.

Table 4.- Welfare Cost: 1974-85.

The results, summarized in table 4, show that the welfare cost from oil crises was very different between countries. Again, southern European countries, with a lax monetary policy, had a larger welfare cost, ranging from 6% of output in Spain to more than 8% in Portugal, so consumers should have been compensated in each period with a significant fraction of GDP in terms of consumption in order to make up for the loss in welfare derived from the different oil crises. The welfare cost in the rest of the countries was moderate (Belgium and Luxembourg) or even small. This gives us an idea of the significant amount of the loss in welfare brought about by the different oil crises for a small open economy, and the importance of managing monetary policy to accommodate such kinds of shocks.

6.- Conclusions

In this paper, we have analyzed the effects of oil price shocks on the business cycle of the EU-15 countries. The model used for this analysis is based on the standard dynamic general equilibrium model for a small open economy in which oil is included as an imported productive input. The price of oil and the interest rate are assumed to be set by international markets. The calibration of the parameters of the model is carried out by taking data from the EU-15 economies during the period 1960-2003.

The results show that oil shocks can account for a significant percentage of GDP fluctuations in many of those countries, but the explanatory power is quite smaller for others. That wide range of variation can be explained by differences in the strength of monetary policies. In addition, the model reproduces the cyclical path of the European economies in periods of oil crisis. Finally, we have shown that the increases in the relative price of oil had a negative effect on welfare, particularly in southern European countries, which are historically associated with a lax monetary policy during oil crisis.

References

1. De Miguel, C., B. Manzano and J. M. Martín-Moreno (2003), ‘Oil price shocks and aggregate fluctuations’, The Energy Journal, 24(2), 47-61.
2. Finn, M. (1995), ‘Variance properties of Solow´s productivity residual and their cyclical implications’, Journal of Economic Dynamics and Control, 19, 1249-1281.
3. Greenwood, J., Z. Hercowitz and G. Huffman (1988), ‘Investment, capacity utilization and the real business cycle’, American Economic Review, 78, 971-987.
4. Hamilton, J. (1983), ‘Oil and the macroeconomy since World War II’, Journal of Political Economy, 91, 228-248.
5. Kim, I. and P. Loungani (1992), ‘The role of energy in real business cycle models’, Journal of Monetary Economics, 29, 173-189.
6. Mendoza, E. G. (1991), ‘Real business cycles in a small open economy’, American Economic Review, 81, 791-818.
7. Mork, K. A. (1989), ‘Oil and the macroeconomy when prices Go Up and Down: An extension of Hamilton´s results’, Journal of Political Economy, 97, 740-744.
8. Mork, K. A. (1994), ‘Business cycles and the oil market’, The Energy Journal, Special issue, The Changing World Oil Market, 15-38.
9. Olson, M. (1988), ‘The productivity slowdown, the oil shocks, and the real cycle’, The Journal of Economic Perspectives, 2, 43-70.
10. Pindyck, R. S. (1979), The Structure of World Energy Demand, Cambridge MA: The MIT Press.