the NiGEM and the OECD Interlink predict a first year impact of a 50% increase in oil prices
on consumer prices of 0.3% to 0.6%, and a cumulated 0.5% to 1.0% impact over 3 years.12
This is significantly higher than the 0.3% we would get for a 50% increase in energy prices
on the HICP excluding unprocessed food and energy after 4 years, but roughly in line with
our results when allowing for an additional 0.5%-1.0% due to the direct effect of energy on
the overall HICP. Looking at the results from small-scale models, Hahn (2003) suggests that a
50% increase in oil prices leads to a 0.9% increase in overall consumer prices after 1 year,
1.6% after 2 years and 2.2% after 3 years, somewhat higher than the results above.
A specificity of our approach is that, unlike other models, we estimate the impact of energy
and non-energy commodities separately and we further split the latter into food and industrial
raw material prices. This is particularly important in our pricing chain analysis, as different
commodity prices might have a different impact on the individual price components.
The results for food commodity prices are shown in Chart 7. The effect is the strongest for
PPI consumer goods (around 0.02% after 16 quarters, with most of the impact coming
through within the first year), while food commodity prices have, unsurprisingly, not been
found to be significant for the other two PPI components.
Chart 7 Impact multiplier of food commodity prices
(deviation from baseline following 1% increase in food commodity prices)
Effect of 1% increase in food comm.
prices on producer prices
(deviation from baseline)
Effect of 1% increase in food comm.
prices on consumer prices
(deviation from baseline)
PPENE: PPI energy; PPINT: PPI intermediate goods; PPCONS: PPI consumer goods; CPFDPR: HICP
processed food; CPNEIG: HICP non-energy industrial goods; CPSERV: HICP services, CPEX: HICP excluding
unprocessed food and energy.
12 See European Central Bank (2004).
ECB ■
Working Paper Series No 1104
November 2009∣ 19