This website contains data and programs for the paper

  James G. MacKinnon, Morten \O. Nielsen, and Matthew D. Webb, "Testing
  for the appropriate level of clustering in linear regression models",
  Journal of Econometrics, 235, 2023, 2027-2056.

  https://www.sciencedirect.com/science/article/pii/S0304407623001021

The working paper version is here:

  https://ideas.repec.org/p/qed/wpaper/1428.html

The website contains the following files:

mnwsvt.ado -- A Stata program for computing both score variance (SV)
              tests and Ibragimov-Mueller (IM) tests.
	      
mnwsvt-readme.txt -- The readme file for mnwsvt.ado.

readme.txt -- this file.

star_test.dta -- The STAR dataset used in the empirical example.

table_maker.do -- A Stata program that calls mnwsvt.ado a number of
                  times to generate the results in Table 1.

Because mnwsvt.ado is quite slow when bootstrapping SV tests, the
bootstrap SV results in the paper were obtained using a Fortran
program rather than mnwsvt.ado. Thus it is impossible to replicate
them exactly. You should be able to get reasonably similar results if
you use B=999. But this will take a lot of time.

Although mnwsvt.ado does compute IM tests reasonably quickly, the
results for those tests in the paper were also obtained using a
Fortran program. Since they depend on random numbers, they cannot be
replicated exactly.