Lex Renda, in his 1991 dissertation, The Polity and the Party System: Connecticut and New Hampshire, 1840–1876, estimated voter transitions in Connecticut elections using Leo Goodman’s 1953 method of ecological regression. Below I compare my estimates to his for the transitions from the 1853 to 1854 elections and from the 1854 to 1855 elections.
Voter Transitions from 1853 to 1854, estimated using multinomial–Dirichlet
Democrat Whig Free Soil Temperance Non-voting share in 1853
Democrat 26.0 0.7 0.2 0.4 4.6 31.9
Whig 0.8 15.4 0.4 3.8 1.0 21.4
Free Soil 0.8 1.5 1.4 4.4 1.2 9.3
Non-voting 1.1 2.2 0.6 2.2 31.3 37.4
share in 1854 28.7 19.8 2.6 10.8 38.1 100.0
Voter Transitions from 1853 to 1854, estimated by Lex Renda using Goodman
Democrat Whig Free Soil Temperance Non-voting share in 1853
Democrat 31.1 -0.5 -1.6 2.7 0.3 32.0
Whig -1.0 19.4 2.1 0.8 0.2 21.5
Free Soil 0.1 1.2 2.9 4.1 0.9 9.3
Non-voting -0.8 0.1 -0.7 3.4 35.2 37.2
share in 1854 29.4 20.2 2.7 11.1 36.6 100.0
Voter Transitions from 1854 to 1855, estimated using multinomial–Dirichlet
Democrat Whig Know Nothing Non-voting share in 1854
Democrat 17.3 1.3 7.8 2.3 28.7
Whig 2.8 4.7 9.8 2.5 19.8
Free Soil 0.2 0.4 1.2 0.8 2.6
Temperance 2.0 1.3 6.3 1.2 10.8
Non-voting 4.6 1.3 2.6 29.6 38.1
share in 1855 26.9 9.0 27.7 36.4 100.0
Voter Transitions from 1854 to 1855, estimated by Lex Renda using Goodman
Democrat Whig Know Nothing Non-voting share in 1854
Democrat 19.1 -2.2 10.5 1.2 28.7
Whig 5.6 7.8 6.6 -0.3 19.7
Free Soil -2.3 1.0 2.5 1.4 2.6
Temperance 2.5 2.9 6.2 -0.8 10.8
Non-voting 2.7 -0.2 2.6 33.1 38.2
share in 1855 27.6 9.3 28.5 34.6 100.0