Yeah good question. To add to the confusion, this paper published PNAS says "2.83 ± 1.0 t of CO2-equivalents per capita". Per capita is different from per household but might be a better metric

It also states that using just national energy statistics (this study actually analyzed 93 million households instead) you'd arrive at 3.19 t CO2 per capita which falls in line with their results

https://www.pnas.org/doi/10.1073/pnas.1922205117