facebook instagram youtube x

Index numbers

Index numbers are a convenient way to show changes in economic variables, including GDP, consumer spending, investment, productivity, and inflation.

    An index is a mathematical device which tracks changes in a number from a starting point - commonly a base year - and is expressed in terms of the number 100.

The main purpose of index numbers is to put data into a form that enables quick and reliable comparisons to be made.

For example, if we find that country X has experienced an increase in its annual national income of $45 billion, there is not much we can learn from this on its own. If, in the previous year, its total national income was $720 billion, then we can clearly see that income has increased. However, if we convert this to an index, and have a starting point (a base year), comparisons can be made quickly without the need to work out the percentage change - index values already do this.

In this case, if $740 billion is a base figure, then the addition of $45 billion means an increase of 6.25% (45/720 x 100).

There is one last step in the process - to convert this to an index, with the base figure of 100. We could also start the base figure at 1000, but 100 is convenient as it allows for the quick interpretation of changes. In this case, we simply add the 6.25 (%) to 100, to give 106.25. Hence, country X has an index value of 106.25, and has grown by 6.25%. For decreases, we use the negative number to reduce the index value by the percentage decline. So, if national income for country X fell by 4%, the index value would be 96.

Using an index to track inflation rates

Price inflation can only sensibly be tracked by using an 'index'. This is because there is no single number which represents all the goods and services that exist and are subject to prices changes.

Index formula

Inflation basket for a family.

For example, if we apply the information we have on the Spanish family, the starting value is €3, and if there is no change in price, the index value would be:

Inflation basket for a family - example.

Let us say that in the second week (week 2) the price of newspapers increases from €3 to €4, and the price of asparagus increases from €4 to €5, but bottled water is cheaper, falling from €2 to €1.

In this case the price level faced by the family has risen to 22/7, which is €3.14.

Inflation basket for a family- example two.

The new index number for the price change is 104.67. Hence, the index value is a convenient way of showing the percentage change in the price level - in this case, inflation is 4.67%. It is worth noting that, unless the Spanish family increases their income over the period, the increase in the price level will make them worse off in 'real' terms.

Conversely, if the price level falls (in week 3), the index will drop below 100. For example, if the basket of 7 goods drops to €2.90, the new index (compared with the base figure), is:

Inflation basket for a family - example three.

Making the index representative

However, there is an important issue to consider - if we simply add up the price changes for a basket, and then create an 'average' this will not provide us with the real picture.

This is because the measurement of inflation needs to have a subjective element in it to reflect the fact that households do not experience inflation in the same way. For our family, if bread and ice creams both increase in price by 10%, this does not mean that there is a 10% impact on their 'experience' of inflation?

Consider their basket of 7 goods, with the individual prices shown in percentage and index form. A simple average is shown below:

Item Price rises
Bread 10%  (index = 110)
Rice 8%  (index = 108)
Newspaper 4%  (index = 104)
Water 4%  (index = 104)
Bananas 0 - (index = 100)
Ice cream 0 - (index = 100)
Asparagus -5% (index = 95)
AVERAGE PRICE RISE = 721/7 = 103 (index value)

However, this does not mean that the family's experience of inflation is that prices have risen by 3%. The mean average is 3%, but what if the family buys a newspaper every day, and only buys asparagus and ice cream as luxuries and only occasionally. How do we solve this?

The subjective experience of inflation really depends (at least in large part) on the percentage of income spent over time, which depends on the price and the frequency of purchases. Look at the table:

Item Income spent %
Bread 30
Rice 25
Newspaper 15
Water 10
Bananas 8
Ice cream 7
Asparagus 5

So, any increases in the price of bread or rice (which account for 55% of spending) will have a greater impact than increases in the price of ice cream and asparagus (which account for just 12% of spending).

Weighting the index

The solution to the problem is to 'weight' the various items according to the income spent on them over a period of time.

The simplest way to add a weight is to multiply the individual inflation rate by the chosen index number (and then divide by the number of weights allocated, as shown in the table:)

Item Income spent % = weights Individual inflation (%) and index Weighted index
Bread 30 10%  (index = 110) 30 x 110 = 3300
Rice 25 8%  (index = 108) 25 x 108 = 2700
Newspaper 15 4%  (index = 104) 15 x 104 = 1560
Water 10 4%  (index = 104) 10 x 104 = 1040
Bananas 8 0 - (index = 100) 8 x 100 = 800
Ice cream 7 0 - (index = 100) 7 x 100 = 700
Asparagus 5 -5% (index = 95) 5 x 95 = 475
TOTALS 100 weights 10575
    Divide the weighted index [10575] by the weights [100] 105.75

When taking income into account, the index of inflation for the family is 105.75, which means that, as a percentage, inflation was 5.75% over the period (and not the mean average of 3%). Hence a weighted index provides a more 'representative' picture of inflation for this family.

Of course, the weights could be reduced 10 in total, or increased to 1000 - the key point is that whatever the total weights are, once they have been used to create the weighted index, the same number is used to create the final index figure.

Key features of a price index

Key features

  1. The use of samples - not all goods and services are included in the basket, only typical ones and not all outlets.

  2. Surveys are conducted to decide which goods and services are included.

  3. The basket can be updated to reflect changes in spending patterns.

  4. Goods are weighted according to the importance of the good or service in the typical family's spending.

  5. Tracking starts with a base year and uses 100 as a starting measure.

  6. However:

  7. The sample may not be fully representative.

  8. The basket could be slow to update.

  9. Sudden changes in price - such as through an oil price shock - may not be reflected in the index.