Table 5.

Mean of estimates of $\alpha$ for DGP (5.3)–(5.4).

	SUR, Efficient $W$	SUR, Diagonal $W$	SUR, Identity $W$
Panel A: True $E_0=2$ and $I_0=0$, Assumed $E_0=2$ and $I_0=0$
Levels, Cumulative	0.0067	0.0068	0.0067
Levels, Daily	0.0068	0.0068	0.0068
Logs, Cumulative, Trim	0.0068	0.0068	0.0071
Logs, Daily, Trim	0.0067	0.0067	0.0071
Logs, Cumulative, No Trim	0.0080	0.0081	0.0098
Logs, Daily, No Trim	0.0091	0.0097	0.0181
Panel B: True $E_0=2$ and $I_0=0$, Assumed $E_0=1$ and $I_0=0$
Levels, Cumulative	0.0134	0.0135	0.0135
Levels, Daily	0.0135	0.0136	0.0136
Logs, Cumulative, Trim	0.0136	0.0137	0.0143
Logs, Daily, Trim	0.0134	0.0134	0.0143
Logs, Cumulative, No Trim	0.0160	0.0162	0.0198
Logs, Daily, No Trim	0.0182	0.0194	0.0352
Panel C: True $E_0=2$ and $I_0=0$, Assumed $E_0=16$ and $I_0=0$
Levels, Cumulative	0.0008	0.0008	0.0008
Levels, Daily	0.0008	0.0008	0.0008
Logs, Cumulative, Trim	0.0008	0.0008	0.0009
Logs, Daily, Trim	0.0008	0.0008	0.0009
Logs, Cumulative, No Trim	0.0010	0.0010	0.0012
Logs, Daily, No Trim	0.0011	0.0012	0.0023

The table presents the mean of the estimates of ${\alpha }_0$ across 500 simulation draws. The true value is $\alpha =0.0067$. The estimates in different columns combine the deaths and reported cases data and use different weighting matrices.