« Reply #3 on: Monday 28 December 15 05:15 GMT (UK) »
I have found people duplicated on US censuses but not in 1840. There was no reason for me to think that the duplications involved any padding of the numbers for political purposes. The earlier censuses were not conducted on the same specific day everywhere in the country so people could move from one place to another and wind up being enumerated twice.
It would take a massive statewide fraud to inflate the population enough to increase the number of representatives in the US Congress. Local fraud of some kind is much more likely. One of my ancestors was involved in such a case. He was a plaintiff in a law suit which alleged fraud by a local enumerator in the 1905 Iowa state census, the goal apparently being to facilitate a change in the liquor laws of the county. You might be able to spot this sort of census inflation at the local level by examining other names in the county to see if they were also duplicated.
Wiltshire: Banks, Taylor
Somerset: Duddridge, Richards, Barnard, Pillinger
Gloucestershire: Barnard, Marsh, Crossman
Bristol: Banks, Duddridge, Barnard
Down: Ennis, McGee
Wicklow: Chapman, Pepper
Wigtownshire: Logan, Conning
Wisconsin: Ennis, Chapman, Logan, Ware
Maine: Ware, Mitchell, Tarr, Davis