Hausman Test Not Running: No Coefficients in Common (r498)

[Prabha Neupane]

Hi everyone,

I'm running into an issue with the Hausman test after estimating a Mean Group (MG) and Pooled Mean Group (PMG) model in Stata. I am trying to compare the two models to see if PMG is preferable to MG, but I keep getting the error: **"no coefficients in common; specify equations(matchlist)"**. Here’s the process I followed:

1. First, I ran the MG model:
```
. xtpmg d.int_rate_deposit ib_bank, lr(l.int_rate_deposit ib_bank) ec(ECT) replace mg
```

This ran fine and produced the expected results.

2. Next, I ran the PMG model:
```
. xtpmg d.int_rate_deposit ib_bank, lr(l.int_rate_deposit ib_bank) ec(ECT) replace pmg
```

However, for the PMG model, while the short-run coefficients are estimated, the long-run coefficient for `ib_bank` returns a missing value (shown as `.`).

3. When I try to run the Hausman test between MG and PMG:
```
. hausman mg pmg, sigmamore
```

I get the error message:
```
no coefficients in common; specify equations(matchlist)
r(498);
```

I'm unsure why the long-run coefficient is missing for PMG and how to resolve the Hausman test issue. Has anyone encountered this problem before or can point me in the right direction?

Thanks in advance!

[Andrew Musau]

Probably, you need:

hausman mg pmg, eq(1:1) sigmamore

Note that xtpmg is from SSC (FAQ Advice #12).

[Prabha Neupane]

Thank you very much, it worked perfectly. I am new to this forum and have just started learning Stata. I will try to adhere to the advice provided in the FAQ from now on.