Base Effect

Definition

The base effect arises not from the absolute change in an indicator but from the percentage conversion performed on it, which divides that change by some base value. The UK Office for National Statistics (ONS) defines it as "the effect that the corresponding base period from the previous year has on current growth rate estimates." When the base is low, the same absolute increase appears as a large growth rate; when the base is high, the same increase appears as a small one.

History

The base effect is not an event but a general property inherent in percentage calculation, yet it became a practical concern in statistical work through debates over base years and inflation indices. In a 2005 HICP study the European Central Bank (ECB) treated it as the contribution of an unusual or extreme change in the price index (or a sub-index) during the base period to the change in the annual inflation rate. That definition seems economically sound but leaves several issues open. In 2021 the ECB described the effect as what happens to a year-on-year inflation rate when an unusually large month-on-month change from twelve months earlier drops out of the index, which meant that in spring 2021 oil and energy prices exerted upward pressure on annual rates even while staying near their low post-pandemic levels. Both ECB documents failed full-text PDF extraction, so these formulations are preserved only at the level of search snippets.

Examples

The simplest illustration is a price index running 100, 150, 200. The rise of 50 from period 1 to period 2 yields a 50 percent increase, but the identical absolute rise of 50 from 150 to 200 yields only 33.33 percent, because the base used for the calculation has changed from 100 to 150.

The pandemic-era figures cited by the ONS show the scale involved. Steep declines in April 2020, retail sales at -18.1 percent, goods exports at -20.6 percent and monthly GDP at -18.7 percent, became the starting point for twelve-month growth rates, so that double-digit growth could be reported without any boom. The ONS explained that even if UK monthly GDP stayed flat from April 2021 onward, the twelve-month growth rate would jump to nearly 25 percent and remain in double digits through July. On prices, fuel and lubricants contributed -0.30 percentage points to the twelve-month CPIH rate in April 2020 but added +0.28 points in April 2021, the two effects together adding 0.58 points to the latest annual inflation rate. Such distortions vanish from annual figures after a few months, but until then users of official statistics must beware of base effects.

Korean cases are documented as well. Industrial production growth for December 2009 was reported at its highest in ten years year on year, largely because the base month, December 2008, fell in the slump of the global financial crisis. The Bank of Korea, using a price example, explains that if agricultural prices had surged the previous summer after typhoon damage, this summer's year-on-year increase in agricultural prices would appear relatively low, whereas if fruit prices had collapsed the previous autumn on oversupply, this autumn's increase would likely appear high.

Relations

The base effect is closely tied to the base year. The base year serves as the reference point that normalizes rates of change, functioning much like the denominator in a comparison. In measuring inflation with price indices, the Paasche and Laspeyres formulas produce different results because of base effects. Percentages in time series that are compounded annually over several years can mislead for the same reason.

Distinctions

The base effect must be distinguished from a genuine change in economic activity. A high base effect misleads because percentages fall even though the absolute gap in the base keeps widening as the measured value or population grows over time; a low base effect misleads because percentages rise even as the absolute gap narrows. Because of the problems base effects create for volatile prices, especially headline inflation, core inflation is used as an additional indicator of inflation developments. Several methodologies adjust for base effects in calculating economic growth, and they are frequently used in financial market analysis. When reading an indicator, the first thing to check is whether the base period contained an exceptional situation such as typhoon damage or oversupply.