Philip investigated the relationship of per capita beer sales to tax rate and income

based on 50 states in America for the period 1975 - 2005. The table below shows the

results of pooled ordinary least square (OLS), fixed effects (FEM), and random

effects (REM). The dependent variable is per capita beer sales.

Variable

OLS

FEM

REM

Constant

1.4192 (24.37)

1.7617 (52.23)

1.7542 (39.22)

Beer tax

-0.0067 (-2.13)

-0.0183 (-9.67)

-0.0181 (-9.69)

Income

-0.00000354 (-1.12)

-0.000020 (-9.17)

-0.000019 (-9.10)

R

2

0.0062

0.0052

0.0052

Notes: Figure in parentheses are the estimated t ratios.

(i)

Do the results make economic sense? Justify your answers.

(4 marks)

(ii)

Is there a vast difference in the results produced by the three models? Justify

your answers based on the estimated coefficient of the beer tax and income

variables.