Beer, along other products has been described as a "defensive", meaning that its consumption, and therefor, its demand, is inelastic.
Several works have estimated the income elasticity for beer, confirming the hypothesis of beer sales being resilient to income. Diverting from this approach, this paper focuses in the analysis of beer consumption in a context of business cycle.
This table presents results of alternative test of a cointegrating relationship among the five variables, comparing with the use or not of lags.
The Engle - Granger technique is using for testing the residuals of an OLS regression of beer on contemporaneous values of the cyclical variable.
The Dickey-Fuller method is used to corroborate the stationarity.
The johansen trace test, is used to validate the procedure for producing cointegrated vectors.
Giving the exitisting cointegration among economic variables, the use of an error-correction methodology is required.
Results.
Residuals are stationary.
Variables are cointegrated.
The contemporaneous elasticity of beer with respect to income is between 0,5 and 0,7.
About beer consumption:
Is pro-cyclical as measured by its responsiveness to industrial production.
Is positively related to unemployment, supporting either the view that it is cheaper alternative to other alcoholic beverages.
Excise tax rates are negatively correlated.
The results shown in these tests reveal that beer consumption is not responsive to the business cycle or any variation regarding a long term equilibrium.
After the tests, there is no evidence to support causality between the business cycle and the pattern of beer sales. Proving that beer is, in fact, inelastic to the forces of markets and economic trends.