What is a Base Period?

A base period is a point in time for which data is gathered and used as a benchmark against economic data from other periods to interpret them on a common basis. Base periods are often used in finance and economics applications, such as measuring inflation or other variables subject to change based on the passage of time.

Base period may also be referred to as "reference period", "basis period", or "index period."

Key Takeaways

  • Base period refers to the benchmark against which economic data from other periods is measured.
  • Comparing other data points to a constant base period allows analysts to spot changes and distinguish long-term trends from short-term fluctuations.
  • The choice a base period can influence an observer's perspective on the data, which is known as the base effect.

Understanding Base Periods

The base period can be thought of as a common yardstick for economic data. Rather than stating each data point in a series as a raw number, it can instead be stated as a proportion or percentage of the data value in the base period. The value for the base period is customarily set as the unit of measure, usually 1 or 100, and all other data points are restated as decimal, fractional, or percent values of the data value for that period.

Comparing each data point to the base period can be a convenient way to handle data series that consist of large or complex numbers. Each data point in the indexed series can then be easily interpreted as the proportion, percent change, or growth rate of the underlying data series over time, relative to the base period.

Example

For example, if a price index has a base year of 1990, and current prices are being compared to prices in that time period to construct a time series index, then the price level in all other years would be stated as a percentage of the 1990's price level. The price index for 1990 might be assigned a value of 100 and price levels for other years would have values proportionally greater (or less than) 100 in proportion to the ratio of the actual price levels of those years.

The calculation of the price level for 1995 might be calculated by taking the proportion:

1990  price level 100 = 1995  price level x \frac{1990 \text{ price level}}{100}=\frac{1995 \text{ price level}}{x} 1001990 price level=x1995 price level

and solving for x:

1990  price level 100 × 1995  price level = x \frac{1990\text{ price level}}{100\times1995 \text{ price level}}={x} 100×1995 price level1990 price level=x

Alternatively, though less commonly, the base period may refer to comparing each data point to a past data value using a constant interval of time rather than a constant base period. This technique does not create a consistent index comparison over time, but can help eliminate the effect of seasonal or short-term fluctuations in data. Year-over-year, or month-over-month comparisons are examples of using past data at a constant interval as a basis for comparison to current data.

The use of base periods to index data is not constrained to financial applications. Many natural sciences also regularly use a base period as part of their analytical processes. For instance, to measure changes in global climate patterns, base years must be established.

Base Period and the Base Effect

When constructing an index, the choice of a basis for comparison can influence how the data can be interpreted and should be chosen carefully to illuminate the desired objective that the data is being used for. Atypical values or abnormal conditions in a base period can lead to comparisons that distort the trends in a data series. This distortion is sometimes called the base effect.

For example, suppose the City of New York institutes new building codes that go into effect on June 1st of a given year. In the month of May, builders scramble to initiate new building projects to avoid the expense of complying with the new codes. This might lead to a situation where the data for building starts show an abnormally high value for May and an abnormally low value for June, as builders move up their construction schedules, which does not reflect any underlying trend in the data, only the one-off regulatory change.

In this case, choosing either the May or June data point as the base year in a comparison or to construct a time series index would lead to wildly distorted results in the resulting transform data since every data point in the index will be comparing present data to an abnormally high (or low) value in the denominator. An analyst would be well advised to choose a more typical value as the base period for the comparison of later data points.