There is a continuing interplay between models and data in the study of economic
relationships. A model is a simplified framework to organize how we think about a problem.
Data or facts are essential for two reasons. They suggest relationships which we should aim
to explain and they allow us to test our hypotheses and to quantify the effects that they
imply.
Tables present data in a form easily understood. Time-series data are values of a given
variable at different points in time. Cross-section data refer to the same point in time but to
different values of the same variable across different people.
Index numbers express data relative to some given base value.
Many index numbers refer to averages of many variables. The retail price index summarizes
changes in the prices of all goods bought by households. It weights the price of each good
by its importance in the budget of a typical household.
The annual percentage change in the retail price index is the usual measure of inflation, the
rate at which prices in general are changing.
Nominal or current price variables refer to values at the prices ruling when the variable was
measured. Real or constant price variables adjust nominal variables for changes in the
general level of prices. They are inflation-adjusted measures.
Scatter diagrams show the relationship between two variables plotted in the diagram. By
fitting a line through these points we summarize the average relationship between the two
variables. Econometrics uses computers to fit average relationships between many
variables simultaneously. In principle this allows us to get round the ‘other things equal’
problem, which always applies in two dimensions.
Analytical diagrams are often useful in building a model. They show relationships between
two variables holding other things equal. If we wish to change one of these other things, we
have to shift the line or curve we have shown in our diagram.
To understand how the economy works we need both theory and facts. We need theory to
know what facts to look for: there are too many facts for the facts alone to tell us the correct
answer. Facts without theory are useless but theory without facts is unsupported assertion.
We need both.
