.comment-link {margin-left:.6em;}

Thesis & Antithesis

A critical perspective on energy, international politics & current affairs

My Photo
Location: Washington, D.C.

greekdefaultwatch@gmail.com Natural gas consultant by day, blogger on the Greek economy by night. Trained as an economist and political scientist. I believe in common sense and in data, and my aim is to offer insight written in language that is clear and convincing.

02 July 2006

How vulnerable to oil prices? Part 2

My post on the observation that high oil prices have not forced a slowdown in economic growth was based a macro view of the world (“How vulnerable to oil prices,” 27 June 06). A micro analysis supports the overall thesis that energy costs are less important than they used to be, and it also reveals trends which are consistent with theory and common sense. To conduct a micro analysis, I asked the question: how are households and industries affected by high oil prices?


Personal consumption is what non-government entities spend every year (the difference between income and consumption is savings / investment). The question to ask in terms of oil prices is this: how much consumption goes to energy-related expenses? The graph below plots this: the blue line calculates total spending on gasoline oil, fuel oil and coal, and electricity and gas. Then, the red line plots what percentage of total consumption these expenditures make up.

As expected, although energy-related expenditures have increased since 1960, they form a smaller share of total consumption; and even though this trend is reversing, the 2005 figure (for first half of 2005) is still lower than either pre-1973 levels and considerably less than the energy crisis years in the 1970s. What this graph shows is that despite high energy prices, the amount that the US spends on energy as a percentage of its total spending is still small, though increasing.


A second important question in examining GDP growth and high energy prices is to look at how different industries are affected by higher oil prices. The hypothesis here is straightforward: industries that rely more on energy should be hardest-hit and we should expect those industries to experience a greater slowdown when compared to less energy intensive industries. The figure below examines that data for this hypothesis.

On the x-axis is a measure of energy intensity measured as follows: money spent on energy inputs over value added. This is a crude measure of energy intensity, but it is a good proxy, and we can expect that industries with higher values rely more on energy for their well-being. The y-axis plots annual growth in value added for 2004. Our hypothesis would say that we would expect a downward sloping line: the more energy intensive, the less industries grew.

This is what the data show: I have circled utilities and transport, both energy intensive and both confirming the hypothesis. I have also circled mining and farms; the former grew despite energy intensity because it includes oil companies which make more money with higher prices, while the latter grew in large part as a recovery from slow (and negative) growth in the past few years. The rest of the industries reveal a similar pattern: more energy intensity equaling less growth.


What is interesting, then, is that even though there is a link between energy intensity and growth, the effect is not large enough to force either a recession or even a slowdown. There are sufficient industries that are not very energy intensive that continue to grow irrespective of high energy prices, thus ensuring that there is no overall slowdown. At the same time, this may be changing, as a greater share of personal consumption is devoted to energy—if prices remain high, the share of consumption devoted to energy may stabilize, and we will probably need to see a further increase in prices to detect a more sizeable effect on economic growth.



Anonymous Meave said...

I'm not seeing much of a correlation here... Utilities & Transport have over three times the energy intensity of the other industries (save for Mining & Farming), yet annual growth is about the same. The plot indicates that energy intensity is unrelated to industry growth -- which is quite interesting.

10:19 AM  
Blogger Nikos Tsafos said...

I think the correlation exists, it just isn't remarkable. Take out mining and farms, and the r-square (an indicator of how much of the variation in one variable can be explained by changes in the other) is about 0.37, which is not bad. Even with utilities and transport there is a nice downward slopping line. In the smaller subgroup (the green "other" category), the R-square reaches around 0.62, which is pretty good.

Still, I think the key to this issue lies in the graph I posted on my third part to this series which shows how energy costs are just not significant enough to affect industry performance.

10:04 PM  

Post a Comment

Links to this post:

Create a Link

<< Home