Presenting Data: Tabular and graphic display of social indicators
 
Gary Klass
Illinois State University
© 2002

Note: The website will be discontinued shortly, to be replaced by the Just Plain Data Analysis site

Home Good Tables Good Charts Divide Analyzing Budgets Election Data Education Poverty References Course page Chart of the week your comments

Divide: Adjusting for inflation and other deflators
(under construction)

Introduction

When constructing a social indicator, there are often several different divisors to choose from in order to standardized the data to account for difference in population and prices.  In the case of many monetary time series indicators, it is often best to present the data constant, inflation adjusted dollars.  Depending on the count, or numerator, of a statistical indicator, certain divisors are often better to use than others or are used for different effect. 

In each of the following time series charts the same numerator (National Defense Spending) is used with different divisors.  Each represents a valid presentation of the data, but depending on which denominator is used one could conclude that National Defense spending is steadily or dramatically increasing, steadily or dramatically decreasing, or holding relatively constant.

Table D-1: Defense spending
in current and constant dollars
Table D-2: Defense spending
as a percent of GDP
Table D-3: Defense spending
as a percent of total spending
Table D-4: Defense spending per capita
in current and constant dollars
Table D-5: Defense and Total spending,
current dollars, 1980=100
excel file

 

Consumer Price Index

The Consumer Price Index is the most commonly used measure of inflation and is an appropriate deflator for indicators of personal income and spending. 

Other Price Indexes

Because the Consumer Price Index is based on the prices of a market basket of goods that consumers typically purchase, it is not a good deflator to use for measure of aggregate government spending.  Often government revenue and spending indicators (either cross national or time series) are deflated as a percentage of gross domestic product.  Note that the the GDP generally increases at a rate faster than inflation; dividing a measure of government spending by GDP will thus result in a lower rate of growth than if and inflation measure were used as the divisor.

The broadest measure of price changes is the GDP (chained) Price Index, also called the GDP implicit price deflator.  The GDP deflator is used to construct the constant dollar measure of GDP by can also be used  

Purchasing power parity

When making cross national comparisons of government revenues, expenditures and benefits or personal income one begins by converting the monetary figures into a common currency.  Cross national tabulations of per capita health or military expenditures, average annual social security benefits, or median family income are examples of commonly reported indicators that are usually represented in US dollars.

Table D-xx: GDP adjusted for Purchasing Power Parity
source: Smeeding and Rainwater, (2002)

Converting to US dollars, however, is usually not sufficient to take into account differences in prices between countries.  Consider the problem: $100 worth of Euros or British pounds  may not buy the same amount of goods in the US as it does in France or Britain (although, in theory government regulations and barriers to trade were eliminated, prices would equalize across countries).  To account for the differences in prices, the OECD reports data on Purchasing Power Parities (PPP). 

Table D-xx reports per capita GDP data for a selected set of OECD nations both in US dollars and adjusted for purchasing power parity.  The United States ranks third in terms of per capita GDP, behind Switzerland and Norway. When one takes into account the substantially higher standard of living in those countries, however, the Untied States' GDP is shown to be substantially higher. 

% of GDP and per capita measures.

Per capita and percent of the population measures are the most common use of a divisor in the construction of social indicators. 

The decennial census is the source of the most accurate counts of the US population (and still these suffer from undercounting) and other annual counts of population are based on estimates conducted by the Bureau of Census.  Because these are generally less accurate in the later years of a decade, some Census data series, such of the poverty rate estimates, are corrected when the new population numbers become known. 

Some variations of per capita measures involve a narrower definition of the population:  The National Center for Health Statistics defines the birth rate as the count of the number of births per 1,000  estimated population; The fertility rate is the same count of births per 1,000 women aged 15 - 44 years of age.

For both cross national and time series presentations, measures of GDP are often used to standardized the governmental revenue and expenditure data.   One problem with "% of GDP" measures is that time series trends often fluctuate more because of changes in the country's GDP than changes in numerator.  Often governmental spending will increase incrementally at a steady rate while measures of spending or taxation as a percent of GDP measures will show dramatic changes, increasing during recessions and decreasing during periods of economic growth. (this is discussed on the budget page).   [In this essay Timothy Noah discusses the implications of using the GDP deflator when analyzing defense spending and suggests that adjusting the spending data for inflation would be a better indicator.]

Indexing to 100

In times series displays, comparing two or more data trends measured on different scales using the same Y-axis, generally results in the trend measured on the smaller scale to appear as a flat line at the bottom of the graph.  Thus in Table D-6, it looks as if national defense spending has been increasing at a rather dramatic rate, while foreign aid has remained relatively flat.  In truth, foreign aid has been increasing at a faster rate than defense spending in recent years (due largely to the latest Bush administration's AIDS initiatives).  This can be seen in table D-7, where both trend lines are indexed to 100 at the base year of 1980.

Table D-6: National Defense
and Foreign Aid Spending, 1980-2004
Table D-7: National Defense
and Foreign Aid Spending, 1980-2004
(1980 = 100)

(Excel file)

Indexing trends to a base year set at 100 has the advantage of representing the cumulative percentage change in the indicator.   The defense spending measure in table D-7 increases from 100 in 1980 to 340 in 2004.  Thus the 2004 level of spending is 340% of the 1980 base.

Note that index the base year does not preclude or obviate the need for using other divisors (see table D-5, above).

Data Sources:

Reference:

1 Robert Sarh, "Using Inflation-Adjusted Dollars in Analyzing Political Developments," PSOnline (4/17/04)  http://www.apsanet.org/PS/april04/sahr.pdf

Timothy M. Smeeding and Lee Rainwater, "Comparing Living Standards Across  Nations: Real Incomes at the Top, the Bottom, and the Middle," SPRC Discussion Paper No. 120
December 2002.  http://www.sprc.unsw.edu.au/dp/DP120.pdf