SUR model of demand for Roll-Your-Own (RYO) tobacco and manufactured cigarettes (MC)
| Ln(RYO) | Ln(MC) | |
|---|---|---|
| Constant | 6.892*** (2.218) | 10.066*** (2.082) |
| Ln(price of RYO) | −0.441** (0.220) | |
| Ln(price of MC) | 0.867** (0.373) | −1.033*** (0.221) |
| Ln(average weekly income) | −0.804* (0.436) | 0.397 (0.456) |
| Non-price policy variable | −0.004 (0.002) | −0.004 (0.002) |
| Adjusted R2 | 0.27 | 0.79 |
| Breusch–Pagan test*† | 0.043 | |
| Portmanteau Q (four and eight lags)*‡ | 0.26; 0.084 | |
| Jarque–Bera test*§ | 0.057 | |
| Total system observations | 168 |
SEs in parentheses; ***p<0.01, **p<0.05, *p<0.1.
*p Values.
†Breusch–Pagan test confirms that at a 5% level the error terms of RYO tobacco and MC demand are correlated and therefore the use of a SUR framework is appropriate.
‡Portmanteau test for serial correlation in the error terms up to lag eight.
§Jarque–Bera test for multivariate normal distributed error terms.
SUR, seemingly unrelated regression.