[BUG] Different standard errors between cluster(clustvar) and vce(cluster clustvar)

[eric-roca]

Before submitting the bug report

1. Verify that you are using the latest versions of reghdfe and ftools (which reghdfe). Note that the latest version is usually on Github and not on SSC.

 which reghdfe
/home/eric/.local/share/ado/plus/r/reghdfe.ado
*! version 6.12.2 02Nov2021

. which ivreghdfe
/home/eric/.local/share/ado/plus/i/ivreghdfe.ado
*! ivreghdfe 1.1.1  14Dec2021 (experimental -margins- support)
*! ivreghdfe 1.1.0  25Feb2021
*! ivreg2 4.1.11  22Nov2019
*! authors cfb & mes
*! see end of file for version comments

3. Verify that your Stata is updated (update query).

. update query
(contacting http://www.stata.com)

Update status
    Last check for updates:  17 Mar 2022
    New update available:    none         (as of 17 Mar 2022)
    Current update level:    15 Feb 2022  (what's new)

Possible actions

    Do nothing; all files are up to date.

Bug report

Please complete the following information:

• Stata version: 17.0
• OS: Linux, kernel 5.16-14

Behavior

• Expected behavior: a clear and concise description of what you expected to happen.
  Obtain the same standard errors using cluster(clustvar) and vce(cluster clustvar) as is the case with other Stata commands.

• Actual behavior: what was the error. If possible, copy-paste output. For instance:
  The standard errors differ.

Steps to reproduce the problem

clear all
set obs 1000
set seed 1111

gen x = rnormal()
gen z = 0.2 * x + 5 * rnormal() 
gen e = rnormal()
gen y = x + e

* Construct some clusters
gen g = mod(_n, 25)

ivreg2 y (x = z), cluster(g)

ivreghdfe y (x = z), vce(cluster g)
ivreghdfe y (x = z), cluster(g)