Ohio-Style Apportionment (temporal variation).

[Boss_Rahm]

Do you have lists of the 23 states where Democrats would be likely to win the extra seat, and 15 states were Republicans would?

This is just going by the House popular vote in each state in 2012.

Democrats win the marginal seat in AL, CA, FL, GA, IL, IN, IA, KY, MD, MA, MI, MN, NH, NM, NY, ND, OK, RI, SC, SD, VT, VA, and WI.

Republicans win the marginal seat in AK, AR, CT, ID, KS, MS, MO, MT, NJ, OH, OR, PA, TX, WA, and WY.

[jimrtex]

Do you have lists of the 23 states where Democrats would be likely to win the extra seat, and 15 states were Republicans would?

This is just going by the House popular vote in each state in 2012.

Democrats win the marginal seat in AL, CA, FL, GA, IL, IN, IA, KY, MD, MA, MI, MN, NH, NM, NY, ND, OK, RI, SC, SD, VT, VA, and WI.

Republicans win the marginal seat in AK, AR, CT, ID, KS, MS, MO, MT, NJ, OH, OR, PA, TX, WA, and WY.

Applying the 2012 results to my prospective assignments for 2012, 2014, 2016, 2018, and 2020 would result in the extra seats going:

2012 12D, 6R
2014 11D, 7R
2016 12D, 6R
2018 11D, 7R
2020 11D, 7R

My concern about 2018 ends up not being true. NH, ND, RI, and VT added a Democrat to the results, while AK, AR, and MT added a Republican.

Overall, there are 57 extra Democratic representatives and 33 Republicans. This might be, as you conjectured, due to a Zipfian distribution leading to smaller states having a greater likelihood of smaller fractions (by the time you get to California the fraction becomes quite uniformly random).


I thought of another approach. Consider a state entitled to 1.29 or 5.29 representatives. Under conventional apportionment, they would have 1 or 5 representatives for an entire decade.

The same would be true if they were entitled to 1.11 or 5.11 representatives.

Under our alternative method, all of these states would get an extra representative in one session over a decade, with the session assigned randomly. We treated all these states as having a fraction of 5.2, which is better than ignoring the fraction entirely.

But let's consider the state entitled to 5.29 representative. Ideally, they should have 5.29 representative every session. But that is impossible. But we can at least give them 26 representatives over the decade, 5 in four sessions, and one in the other. Which session should that be.

Through one session, they should have 5.29 representatives. Apportioning five representatives to the state is fair.

But through two sessions they should have 10.58 representatives which is closer to 11 than 10. So they should get the extra representative in the second session.

We can think of rounding down as producing a surplus, and rounding up as producing a deficit which is carried forward to the next session.

So,

2012, 5.29 equals 5 representatives and a surplus of 0.29.
2014, 5.29+0.29 = 5.58 equals 6 representatives and a deficit of 0.42
2016, 5.29-0.42 = 4.87 equals 5 representatives and a deficit of 0.13
2018, 5.29-0.13 = 5.16 equals 5 representatives and a surplus of 0.16
2020, 5.29+0.16 = 5.45 equals 5 representatives and a surplus of 0.45

They almost received an extra representative in the 5th session. There is no reason that the surpluses and deficit could not be carried forward to the next decade.

Meanwhile, for our state entitled to 5.11 representatives, it would not be until the 5th session that the surplus exceeded 0.5 (0.55).

States with large fractions (over 0.5) would receive an extra representative in the first session, but not necessarily all of the earliest session. This might have been the reasoning in the Ohio Constitution of 1851, giving counties with a fraction of 0.6 and 0.8 the earliest sessions, counties with a fraction of 0.2 the 5th session, and those with a fraction of 0.4 the 3rd and 4th sessions.

Note this may be the fairest way to handle the electoral college since it is based on actual population, and eliminates the random assignment of electors.

Assuming we want the same number of total representatives in each session, we should apportion 677, 677*2, ..., 677*5 cumulative representatives using Webster's method from the first through fifth session. Then the apportionment for any session is the difference between the cumulative apportionment for that session, and the previous session. This should never vary by more one.

If we are indifferent to a slight variation in the size of the House, we may use independent rounding. If that is the case, we do not need to use a whole number for the nominal size of the House. 677.432 (or whatever) works equally well.

[Boss_Rahm]

Do you have lists of the 23 states where Democrats would be likely to win the extra seat, and 15 states were Republicans would?

This is just going by the House popular vote in each state in 2012.

Democrats win the marginal seat in AL, CA, FL, GA, IL, IN, IA, KY, MD, MA, MI, MN, NH, NM, NY, ND, OK, RI, SC, SD, VT, VA, and WI.

Republicans win the marginal seat in AK, AR, CT, ID, KS, MS, MO, MT, NJ, OH, OR, PA, TX, WA, and WY.

Applying the 2012 results to my prospective assignments for 2012, 2014, 2016, 2018, and 2020 would result in the extra seats going:

2012 12D, 6R
2014 11D, 7R
2016 12D, 6R
2018 11D, 7R
2020 11D, 7R

My concern about 2018 ends up not being true. NH, ND, RI, and VT added a Democrat to the results, while AK, AR, and MT added a Republican.

I can give you the actual results for the 18 extra seats, using your allocation:

2012: 12D, 6R
2014: 11D, 7R
2016: 10D, 8R
2018: 14D, 4R
2020: 11D, 7R

If you're curious, the 4 states that add a Republican in 2018 are AR, MO, MT, and VA.

[jimrtex]

This is a modernized version of the Ohio method which better complies with current OMOV standards.



The fractional representatives are assigned based on Webster's method. In the Ohio 1851 method, fractions for larger counties were truncated to the lower 0.2 representatives.

Counties and multi-county districts were required to be entitled to 0.8 representatives to have representation. These counties and districts would be entitled to a whole representative. Geauga, Ashtabula, and Columbiana benefited from this bonus.

Under the 1851 Constitution, counties with more than 0.5 of the quota would be entitled to their own district. Smaller counties were combined in pairs, with one trio. The 1851 provided that these multi-county districts be long term, only changing if a county became entitled to its own representative, or fell below the 0.5 threshold.

Rather than trying to simulate a historical process going back to the 1850 Census, I started with all counties having a district, then sequentially eliminated the smallest districts, combining the county with its least populous neighboring district. This could result in the elimination of a small multi-county district, and assignment of the individual counties to different districts. This process continued until all multi-county districts had at least 0.8 of a quota. I then shifted some counties to better balance district populations.

I then applied Webster's method to all the multi-county districts and single-member districts alike. This resulted in many of the multi-county districts having fractional representation. This is not really a problem.

I would likely coordinate the sessions for the extra representatives to maintain regional balance.

Cincinnati (14): Clermont 4 + Butler 1 + (Hamilton, Warren)
Dayton (8): Montgomery 2 + Greene 2 + Clark 1 + (Miami)
Northwest (5): Fulton 2 + Allen 1 + Auglaize 1 + Darke 1
Toledo (6): Lucas 3 + Wood 1 + Hancock 1
Columbus (17): Delaware 4 + Union 1; Fairfield 2 + Licking 2 + Franklin 1
Southeast (8): Pickaway 1 + Ross 1 + Scioto 1 + Athens 1 + Belmont 1 + (Brown, Lawrence)
North Central (5): Erie 2 + Marion 2 + Huron + (Richland)
Northeast (37): Trumbull 3 + Portage 2 + Summit 3 + Cuyahoga 2 + Lorain 3 + Medina 3 + Stark 1 + Tuscarawas 1 + Muskingum 2 + (Lake, Ashtabula, Geauga, Mahoning, Columbiana, Jefferson, Wayne)

Coordination of Northeast would be somewhat complex.

Trumbull 3 and Portage 2 complement each other.
Portage 2 and Summit 3 complement each other. This will lock Trumbull and Summit in phase.
Summit 3 and Cuyahoga 2 complement each other.
Cuyahoga 2 and Lorain 3 complement each other,

Thus Trumbull, Summit, and Lorain will be locked in phase.
Thus Portage and Cuyahoga will be locked in phase.

Two of the Medina sessions will complement Summit and Lorain. The third will overlap one of the three sessions.
This 3rd Medina session will be grouped with Stark, Tuscarawas, and Muskingum.

88 seats would be fixed, with 12 floating.

50.2 representatives would be elected from districts electing more than 2 representatives. 49.8 representatives would be elected from districts electing two or fewer representation.

The smaller districts do not provide much opportunity for proportional representation. The larger counties could, but this would be inconsistent statewide. It would better to divide the larger counties into smaller districts. Larger cities such as Cleveland, Akron, Toledo, Dayton, and Cincinnati might also be divided.

Nine counties would be divided in districts with magnitude between 1 and 2.

Hamilton (4 to 7)
Butler (2 to 3)
Montgomery (3 to 4)
Toledo (2 to 3)
Franklin (6 to 11)
Stark (2 to 3)
Lorain (2)
Summit (3 to 4)
Cuyahoga (6 to 10)

If the goal is to have generally single-county representation, 100 representatives for 88 counties is too few. Basing the size of the legislature on the cube root rule would result in most counties being placed in one or two-county districts.

[Boss_Rahm]

Interesting. I don't know much about what standards courts use for OMOV in state legislatures, but offhand this system seems reasonable. If you increased the minimum threshold to 0.9 quotas, then the maximum deviation for any district would be (State Population)/(10 * (Number of Seats)).

[jimrtex]

Interesting. I don't know much about what standards courts use for OMOV in state legislatures, but offhand this system seems reasonable. If you increased the minimum threshold to 0.9 quotas, then the maximum deviation for any district would be (State Population)/(10 * (Number of Seats)).

The general rule is a 10% range (it should be +/- 5%, or even use a statistical measure such as standard deviation but we are dealing with lawyers and judges).

This is actually a burden shifting threshold. Inside this range the plan is presumed to be lawful, and the burden is shifted to the plaintiff. This was the case in the Arizona legislative redistricting case for the 2010 plan. Democratic seats were consistently underpopulated so that there was a political bias. But Arizona argued that they did so to address VRA concerns (e.g. they would interpret the Gingles test to apply to a compact area with 95 or 96% of the ideal population).

If it is over 10% then the burden shifts to the state. Ohio would argue that they had an interest in maintaining whole-county districts.

One point in the favor is that the counties with over 1% of the population are entitled to 72.438 representatives and they received 72.0.

Deviation is a normal distribution with 60% of the representation between -2% and + 2%, 83% are within -5% and +5%.

OTOH, I started out using a quota of 0.9, and because Geauga and Ashtabula were both between 0.8 and 0.9, I didn't want to merge them in a district with 1.6 representatives. So in a sense I was tuning my statute to a particular set of circumstances.


The other county below 0.9 is Columbiana. I combine it with Carroll that would fix that, but that would carry on through 4 other districts, but that is really not so bad.

I had considered making the law that when a multi-county district elected two representatives, they would be elected by subdistricts. So in this instance Geauga and Ashtabula would have their own representative 3 sessions, and share the representative the other two. Perhaps there might be a requirement that the smaller subdistrict have 40% of the population.

I think I have talked myself into a 0.9 rule. I'll look at potential subdistricts.

[jimrtex]

This is an updated version, which requires a district to have at least 0.9 quotas.



The main change is that Ashtabula and Geauga are combined in a single district, reducing a total of 2.0 representatives to 1.6 districts.

In addition, two counties were added to a Columbiana district, increasing its apportionment by 0.2. This triggered changes in four other districts. The Muskingum and Athens districts each shedded a county, resulting in a loss of 0.4 and 0.2, respectively.

These changes produced a net reduction of 0.8 representatives.

These reductions were compensated by 0.2 additions to Cuyahoga, Montgomery, Licking, and Huron.

Under this plan 85% of the population is in districts with deviation of less than +/- 5%. 65% are within +/- 2%

This requires rearrangement of the extra sessions:

Cincinnati (14): Clermont 4 + Butler 1 + (Hamilton, Warren)
Dayton (7): Montgomery 3 + Greene 2 + (Miami)
Northwest (5): Fulton 2 + Allen 1 + Auglaize 1 + Darke 1
Toledo (6): Lucas 3 + Wood 1 + Hancock 1
Columbus (16): Delaware 4 + Franklin 1; Licking 3 + Fairfield 2
South Central (8): Union 1 + Clark 1 + Pickaway 1 + Ross 1 + Scioto 1 + (Brown, Lawrence)
North Central (5.2): Erie 2 + Marion 2 + Huron 2 + (Richland)
East (10.8): Stark 1 + Columbiana 1 + Tuscarawas 1 + Belmont 1 + (Wayne, Muskingum, Athens, Washington)
Northeast (5.2): Ashtabula 3 + Trumbull 3 + (Mahoning)
Akron (6): Summit 3 + Portage 2
Cleveland (16.8) Cuyahoga 3 + Lorain 3 + Medina 3 + (Lake)

Cleveland is one session short of 10 sessions.
North Central has one extra session beyond 5 sessions.

The one session that the Cleveland are does not have 2 extra representatives, the North Central region will have two extra representative.

Northeast has one extra session beyond five sessions.
East is one session short of five sessions.

The one session that the Northeast has two extra sessions, the East region will not have an extra session.

In the previous message, I had suggested multi-county districts might be subdivided into two subdistricts for sessions when the district had an extra representative.

Possible splits:

Defiance, Henry, Paulding (52% of population); Fulton, Williams (48%)
Auglaize, Hardin (53%), Mercer, Van Wert (47%)
Logan, Champaign (59%), Union (41%)
Seneca,Wyandot (50%), Hancock (50%)
Sandusky, Ottawa (57%), Huron (43%)
Knox, Morrow (60%), Marion (40%)
Madison, Fayette (56%), Pickaway (44%)
Scioto (58%), Pike, Adams (42%)
Ross (54%), Hocking, Meigs, Vinton (46%)
Belmont (51%), Jefferson (49%)
Ashtabula (51%), Geauga (49%).

In three districts there is a clearly dominant county:

Allen (75%), Putnam (25%)
Columbiana (79%), Carrollton, Harrison (21%)
Tuscarawas (68%), Holmes (32%)

In two districts, the largest county is the middle of a string of three.

Darke (36%), Shelby, Preble (64%)
Huron (38%), Ashland, Crawford (62%).

[jimrtex]

This shows an alternative version of what Ohio might have adopted in 1851.



The 1851 alternative constitution provided for a unicameral legislature elected by proportional representation (the rather odd method of STV used in Cambridge, MA was copied from Cincinnati and is sometimes known as the Cincinnati method. Massachusetts subsequently eliminated the use of STV for town and city governments. Cambridge is grandfathered in, but no other towns or cities can opt-in without a special state law, and Cambridge is stuck with the old system of counting).

The size of the legislature is governed by the cube root rule, resulting in an 126 member legislature.

The minimum magnitude of a district is 3.0. A county with a population of more than 3.0 will have its own districts. Based on the 1850 Census, only Hamilton (Cincinnati) and Cuyahoga (Cleveland) qualified.

The other counties are grouped in multi-county groups with between 3.0 and 6.0 representatives. Most of the groups in 1850 were towards the lower end, but the alternative constitution provided for stability in districts. As long as districts remained between the limits, they would not be changed.

An exception would be made if a county qualified for its own district, in which case the remnant of the district would be treated as a new district. This new district might be smaller than 3.0, and need to be attached to another district (or districts).

Other districts might exceed 6.0 and need to be divided.

I've searched the archives and have been unable to find the division of Hamilton County. If Cincinnati had its own district, the split would have been 7.0 to 2.8. Cincinnati at the time was about four square miles. There would have been a river district along the river, and the newer settled areas to the north would have been combined with the rural areas of the county.

[jimrtex]

This shows groups for the extra seats.



A district entitled to 3.6 representatives will have at least 3 representatives in each of the 5 sessions (1852, 1854, 1856, 1858, 1860), and will have a fourth representative in three of the sessions. That is, the average representation will be 3.6 representatives.

There are five ways the extra sessions may be assigned that minimize clustering.

XXOXO
OXXOX
XOXXO
OXOXX
XOXOX

The 2nd pattern is a rotation of the first pattern by one session to the right.
The 3rd pattern is a rotation of the first pattern by two sessions to the right.
and so on.

These patterns ensure that the three extra sessions are not all consecutive AND that there are not two consecutive sessions with just the base representation. That is the extra representation is spread out throughout the decade,

A district entitled to 3.4 extra representative will have at least 3 representatives in each of the 5 sessions (1852, 1854, 1856, 1858, 1860), and will have a fourth representative in two of the sessions. That is, the average representation will be 3.4 representatives.

There are five ways the extra sessions may be assigned that minimize clustering.

OOXOX
XOOXO
OXOOX
XOXOO
OXOXO

These patterns ensure that the two extra sessions are not consecutive AND that there are not three consecutive sessions with just the base representation. That is the extra representation is spread out throughout the decade.

For every "0.6" pattern there is a complementary "0.4" pattern. If there are nearby districts, one entitled to 0.6 extra representatives and one entitled to 0.4 representatives, they can be assigned complementary patterns, such that an extra representative is assigned to a district in a general area every session.

The above map shows three such pairings:

Belmont 0.6, Trumbull 0.4
Stark 0.6, Summit 0.4
Erie 0.6, Seneca 0.4

A variant of the above pattern is where a district with a "0.6" pattern is grouped with two districts with "0.2" patterns. We can modify the 6-4 pattern to a 6-2-2 pattern by assigning the two sessions of the "0.4" pattern to the "0.2" districts. The order of the sessions for the "0.2" districts can be assigned randomly.

The above map shows one such trio:

Wayne 0.6, Lorain 0.2, Richland 0.2

There are five ways that four extra sessions can be assigned to a district with a fraction of 0.4.

XXXXO
OXXXX
XOXXX
XXOXX
XXXOX
XXXXO

A district with a remainder of 0.2 can be assigned the unused session.

The above map has two examples of such pairings:

Licking 0.8, Knox 0.2
Montgomery 0.8, Champaign 0.2

A trio of districts with remainders of 0.4, 0.4, and 0.2 can be arranged as:

ABABC
CABAB
BCABA
ABCAB
BABCA

Examples in the above map are:

Darke 0.4, Allen 0.4, Hancock 0.2
Franklin 0.4, Ross 0.4, Fairfield 0.2

It is also possible to have groups formed from one 0.4 district, and three 0.2 districts; or five 0.2 districts, though none are present in the 1850 map.

Sometimes more complex combinations are needed.

In the 1850 map Athens 0.8, Muskingum 0.6, and Washington 0.6 share 2 floating representatives. Muskingum and Washington can be arranged so that there is only one session where both districts are represented, while in the other four sessions one or the other is represented, and further the session where both are represented corresponds to the one session where Athens is not.

AAAAB
CBCBC

There are five rotations of the overall pattern and B and C can be swapped (randomly assigned to Muskingum or Washington).

The most complicated grouping is in southwestern Ohio,  where Hamilton 0.8, Clermont 0.6,  Highland 0.6, Warren 0,6, and Butler 0.4 share 3 floating representatives.

These can be rearranged as Hamilton 0.8 and Butler 0.4 electing 1.2 representatives, and Clermont 0.6, Highland 0.6, and Warren 0.6, electing 1.8 representatives. In four sessions the western pair will elect one extra representative, and in one session they will elect two extra representatives. The eastern trio will elect two extra representatives in four sessions, and in one session they will elect one.

The session where the western pair elects two extra representatives will coincide with the session where the eastern trio elects one extra representative.

Ignoring rotations, there are two ways that the 0.8 and 0.4 patterns can be arranged relative to each other. The 0.4 pattern must fill the hole in the 0.8 pattern, and the extra session for the 0.4 district must be two session or three sessions earlier. In these patterns, the third row indicates the session with two extra representatives.

XXXXO
OXOOX
O+OOO

or

XXXXO
OOXOX
OO+OO

There is one way to arrange the 0.6 patterns such that they have one session with one extra representative. The fourth row indicates the session with one extra representative.

XXOXO
OXXOX
XOXXO
OOOO-

To assign all five patterns we first determine which session Hamilton will have nine representatives. We will determine by lot whether Butler will have an extra representative in that session AND the session two or three earlier (in a circular fashion).

We will randomly assign Clermont, Highlands, and Warren to the three patterns of the collective pattern and then rotate them together such that the session with one extra representative coincides with the session that the western trio has an extra representative.


Based on these assignments various regions of the state will have constant representation in every session:

Northwest (10 representatives, 1 floating) Darke 3.4, Allen 3.4, Hancock 3.2.
North Central West (7, 1) Erie 3.6, Seneca 3.4
North Central East (10, 1) Wayne 3.6, Lorain 3.2, Richland 3.2.
Northeast Lake (7, 0) Ashtabula 4.0, Cuyahoga 3.0
Northeast West (7, 1) Stark 3.6, Summit 3.4
Northeast East (11, 1) Columbiana 4.0, Belmont 3.6, Trumbull 3.4
East Central (7, 0) Tuscarawas 4.0, Coshocton 3.0
Central North (7, 1) Licking 3.8, Knox 3.2
Central South (10, 1) Franklin 3.4, Ross 3.4, Fairfield 3.2
Southeast (16, 2) Muskingum 4.6, Athens 3.8, Washington 3.6
Southwest South (24, 3) Hamilton 9.8, Clermont 3.6, Highlands 3.6, Warren 3.6, Butler 3.4
Southwest North (10, 1) Montgomery 3.8, Champaign 3.2, Miami 3.0

Of the 126 representative total, 113 are assigned permanently, and 13 float in 12 regions. 8 regions have one floating representative, 1 region has two floaters, 1 region has three floaters, and two regions have none).

[jimrtex]

We now may assign the sessions in which districts get an additional representative. There are four districts which will have an extra representative in four sessions. While the possibility of a Cincinnati-Dayton-Newark-Athens cabal may be unlikely we want to avoid all four districts being in sync. Not only do we not want all four districts to have an extra representative in the same sessions, we want to avoid a session where all are missing.

We have two bowls. One bowl contains five balls, each with a slip inside reading: Athens, Hamilton, Licking, Montgomery, and None. The other bowl contains five balls, each with a slip reading XXXXO, OXXXX, XOXXX, XXOXX, and XXXOX. We draw one ball from each vase resulting in the following assignments.

1852	1854	1856	1858	1860
Licking	Licking	Licking	Licking
	Montgomery	Montgomery	Montgomery	Montgomery
Hamilton		Hamilton	Hamilton	Hamilton
Athens	Athens		Athens	Athens

We now assign the sessions for the other districts in an area. Knox is assigned the 5th session and Champaign the 1st, filling the holes left by Licking and Montgomery, respectively.

Butler is assigned to the 2nd session hole left by Hamilton. For its other session, Butler may have the 4th or 5th (1st or 3rd would result in two consecutive sessions for Butler). There is one way to arrange the patterns for Clermont, Highlands, and Warren:

XXOXO
OXXOX
XOXOX

This combination results in two representative in every session but the 4th, which matches the session in which Hamilton and Butler both have a representative. Warren, Clermont, and Highland are drawn randomly to these patterns.

There is one way to arrange the patterns for Muskingum and Washington.

OXXOX
XOXXO

This comination results in two extra representatives in the 3rd session, which matches the skipped session for Athens. Washington and Muskingum are drawn randomly to fill these two patterns.

1852	1854	1856	1858	1860
Licking	Licking	Licking	Licking	Knox
Champaign	Montgomery	Montgomery	Montgomery	Montgomery
Hamilton	Butler	Hamilton	Hamilton	Hamilton
Warren	Warren	Highland	Butler	Highland
Highland	Clermont	Clermont	Warren	Clermont
Athens	Athens	Muskingum	Athens	Athens
Muskingum	Washington	Washington	Muskingum	Washington

Similar to the case with the 0.8 patterns, we want to distribute the five 0.6 patterns evenly among the districts. We have already assigned patterns to five districts linked to 0.8 districts.

1: XXOXO Warren
2: OXXOX Clermont, Washington
3: XOXXO Montgomery
4: OXOXX
5: XOXOX Highland

There are four remaining districts which may independently assigned to patterns: Belmont, Erie, Stark, and Wayne. With 9 districts and 5 patterns, 4 patterns will be used twice, and one will be used once. We draw to determine that Pattern 3 will be used once. We then randomly assign the districts to the remaining four patterns (two districts will be assigned Pattern 4). Belmont - 4, Erie - 5, Stark 4, Wayne 1.

1: XXOXO Warren, Wayne
2: OXXOX Clermont, Washington
3: XOXXO Montgomery, (none)
4: OXOXX Belmont, Stark
5: XOXOX Highland, Erie

Our updated table is:

1852	1854	1856	1858	1860
Licking	Licking	Licking	Licking	Knox
Champaign	Montgomery	Montgomery	Montgomery	Montgomery
Hamilton	Butler	Hamilton	Hamilton	Hamilton
Warren	Warren	Highland	Butler	Highland
Highland	Clermont	Clermont	Warren	Clermont
Athens	Athens	Muskingum	Athens	Athens
Muskingum	Washington	Washington	Muskingum	Washington
	Belmont		Belmont	Belmont
Erie		Erie		Erie
	Stark		Stark	Stark
Wayne	Wayne		Wayne

We then fill the complementary patterns: Belmont:Trumbull, Erie:Seneca, Stark:Summit, and Wayne:Richland+Lorain. We draw to determine which sessions Richland and Lorain will have an extra representative.

1852	1854	1856	1858	1860
Licking	Licking	Licking	Licking	Knox
Champaign	Montgomery	Montgomery	Montgomery	Montgomery
Hamilton	Butler	Hamilton	Hamilton	Hamilton
Warren	Warren	Highland	Butler	Highland
Highland	Clermont	Clermont	Warren	Clermont
Athens	Athens	Muskingum	Athens	Athens
Muskingum	Washington	Washington	Muskingum	Washington
Trumbull	Belmont	Trumbull	Belmont	Belmont
Erie	Seneca	Erie	Seneca	Erie
Summit	Stark	Summit	Stark	Stark
Wayne	Wayne	Richland	Wayne	Lorain

There are a final two regions that need to have sessions assigne: Allen 2-Darke 2-Hancock 1; and Franklin 2-Ross 2-Fairfield.

As for the districts with more extra sessions, we will attempt to balance the districts with two extra sessions among all five patterns.

1: XOXOO Trumbull, Summit
2: OXOXO Butler, Seneca
3: OOXOX
4: XOOXO
5: OXOOX

So we wish to avoid assigning additional districts to patterns 1 and 2, and assign the last four districts among patterns 3, 4, and 5. When two districts in a region each have two extra sessions, they must be assigned consecutive-pattern. In this case using 3 and 4; and 4 and 5 will result in the unused patterns being used.

We draw to determine which region will use which pair of patterns, and which of the two districts in region will take which pattern.

Allen and Darke are assigned patterns 3 and 4, with Allen taking 3, and Darke, 4. This leaves patterns 4 and 5 to Franklin and Ross, with Franklin taking 4, and Ross, 5.

1: XOXOO Trumbull, Summit
2: OXOXO Butler, Seneca
3: OOXOX Allen
4: XOOXO Darke, Franklin
5: OXOOX Ross

Hancock and Fairfield receive the unused session in their respective regions.

1852	1854	1856	1858	1860
Licking	Licking	Licking	Licking	Knox
Champaign	Montgomery	Montgomery	Montgomery	Montgomery
Hamilton	Butler	Hamilton	Hamilton	Hamilton
Warren	Warren	Highland	Butler	Highland
Highland	Clermont	Clermont	Warren	Clermont
Athens	Athens	Muskingum	Athens	Athens
Muskingum	Washington	Washington	Muskingum	Washington
Trumbull	Belmont	Trumbull	Belmont	Belmont
Erie	Seneca	Erie	Seneca	Erie
Summit	Stark	Summit	Stark	Stark
Wayne	Wayne	Richland	Wayne	Lorain
Darke	Hancock	Allen	Darke	Allen
Franklin	Ross	Fairfield	Franklin	Ross

The above represents our final session assignments. As a sanity check we check whether the 6 districts with a single session are clustered unevenly.

1: Champaign
2: Hancock
3: Richland, Fairfield
4:
5: Knox, Lorain

This is not quite optimum since session 4 is skipped. However, since these districts are represented infrequently, only once in five sessions over the decade, this is of minimal concern. It is more significant when clustering occurs in four sessions. Thus in assigning sessions, priority will be given to balancing the distributions of districts with four and three extra sessions.

Note the alternative 1851 Ohio Constitution says only that the Redistricting Commission shall assign the sessions in an equitable fashion. They adopted this method, and have followed it ever since.san

[jimrtex]

This illustrates the apportionment by session, 1852-1860.



The actual assignment method may be too complex for the ordinary layman to understand. But it could be explained that the Franklin-Jackson district is entitled to about 3-1/2 representatives, and rather than giving it three representative all of the time, or four representative all of the time, it is given three representatives some sessions and four others so it averages out to 3.4.

Before a particular election only a single map might be published in newspapers showing how many representatives would be elected that year. After the election, the elected representatives for each district along with their party might be shown.

Note: A larger image can be seen by right-clicking on the image and selecting "Open Image in New Tab" (this works in Firefox, other browsers will have something similar).

[jimrtex]

This is the 1860 apportionment for Ohio.



Growth during the decade resulted in the expansion of the legislature according to the cube root rule from 126 to 133 members.

Noble became the 88th County in 1851, when it was formed from Guernsey, Morgan, Monroe, and Washington counties. In real life when this happened, the senate district boundaries permanently divided Noble County.

In this alternate version, the 1850 districts continued to be used throughout the 1850's (1852-1860), because the apportionment was based on the 1850 county boundaries and populations. Voters in the newly established Noble County continued to vote for representative in the Tuscarawas, Muskingum, or Washington districts.

When reapportionment occurred in 1861, the entirety of Noble County was assigned to neighboring district with the smallest population, which was Tuscarawas.

No counties other than Hamilton and Cuyahoga had a magnitude above 3.0 so they remained the only counties entitled to their own district(s). Montgomery (Dayton) and Franklin (Columbus) are close and may qualify in 1870.

No districts exceeded 6.0 magnitude and thus needed to be divided.

Three districts fell below 3.0 and were to be eliminated with the constituent counties assigned to other districts: Coshocton-Holmes; Knox-Morrow; and Fairfield-Perry.

Ohio gained 18.1% between 1850 and 1860. This meant that to retain representation a district had to increase by 11.9%. The enlarged legislature reduced the risk of loss of representation but does not eliminate it.

These three districts were decomposed in inverse magnitude order. Assignment of Holmes to Knox-Morrow brought that district above 3.0 and thus continued in its expanded form. Coshocton was assigned to the Muksingum district.

Fairfield was assigned to Ross and Perry was assigned to Licking.

Lucas supplanted Erie as the largest county in its district, likely due to the development of Toledo as a lake port for the growing northwestern part of the state, which was the last settled area of Ohio.

Meigs supplanted Athens as the largest county in its district. It appears that there was growth along the Ohio River. Prior to the development of the railroads, the Ohio River and Lake Erie were the primary transportation corridors. Travel by horseback, wagon, or carriage was slow.

Hamilton County might have been subdivided into three districts, two within Cincinnati entitled to a total of 9.0 representatives, and the remainder of the county entitled to 3.2 representatives.


Cincinnati and Cleveland had robust growth, and smaller population centers were beginning to differentiate.

Largest cities were

Cincinnati 161K (Cincinnati was the 7th largest city in the USA, a drop of one position from 1850, as it was surpassed by Brooklyn, NY)
Cleveland 43K
Dayton 20K
Columbus 19K
Mill Creek Township, Hamilton County 14K (this has mostly been annexed into Cincinnati, or incorporated as Norwood, St. Bernard, or Elmwood Place).
Toledo 14K

Other towns with more than 5K are Chillicothe, Hamilton, Portsmouth, Salisbury Township Meigs County, Sandusky, Springfield, Steubenville, and Zanesville.

There may be a shadow effect around Cleveland and Columbus as more entrepreneurial persons were drawn in to exchange in mercantile activities, or craftsmen or laborers were drawn to jobs that mostly occurred in cities.


Of persons born in other States but living in Ohio:

Pennsylvania 36%
Virginia 16% (this would include modern West Virginia)
New York 16%
Maryland 6%
New Jersey 4%

Of persons born in foreign countries:

German states: 51%
Ireland 23%
England 10%
France 4%
Switzerland 3%

[jimrtex]

In the 1860's there are 11 floating representatives (of a total of 133). Districts are groups in geographical areas. One representative position floats in each of seven areas. Two representatives float in each of two areas.




As in the 1850's the extra positions are assigned in a way that is randomized with the goal of preventing a district from having (or not having) extra representatives in consecutive sessions, to the extent possible. and to also avoid districts with the same number of extra sessions having those extra sessions in sync.

1862	1864	1866	1868	1870
Hancock	Hancock	Hancock	Richland	Hancock
Licking	Seneca	Licking	Seneca	Licking
Highland	Hamilton	Highland	Warren	Highland
Clermont	Warren	Clermont	Clermont	Warren
Champaign	Franklin	Franklin	Champaign	Franklin
Ashtabula	Ashtabula	Cuyahoga	Ashtabula	Cuyahoga
Washington	Meigs	Washington	Washington	Meigs
Lucas	Allen	Allen	Lucas	Allen
Trumbull	Stark	Columbiana	Stark	Columbiana
Tuscarawas	Muskingum	Tuscarawas	Belmont	Wayne
Miami	Montgomery	Butler	Darke	Montgomery

This is the presentation used for the general public. At a glance they can find their district, and the number of representatives elected in each session. There would be an inset showing the three districts in Hamilton County.

[jimrtex]

This is the 1870 Apportionment for Ohio



Growth during the decade resulted in the expansion of the legislature according to the cube root rule from 133 to 139 members. Statewide growth was 13.9%. To maintain its level of representation a district would need to increase by 9.0%.

I discovered that I had been using Adams' method in apportioning the session. Adams' method is more favorable than Webster's method to less populous districts. The effect in this case is less significant than one might expect since the minimum number of units being apportioned is 15 (3 representatives for five sessions).

In 1860 there would have been no difference. In 1850, one additional session would have been apportioned to Hamilton. In 1870, there were three changes (three districts: Hamilton, Cuyahoga, and Greene gained one session; and three districts: Franklin, Knox, and Summit lost one session.

The 1850 map was based on a schedule in the alternative 1851 Constitution. The constitution does actually specify the method of apportionment, but rather says that the apportionment should be based on population. In this case, the redistricting commission followed the advice of my ggg-grandfather who went by the pseudonym of jimrohio.


Montgomery (Dayton) and Franklin (Columbus) surpassed the 3.0 threshold and were provided their own district. Madison and Greene counties were detached to and appended to multi-county districts. In the case of Greene, it became the largest county in the district formally led by Warren.

Two districts surpassed 6.0 representatives and were divided. The Allen district which had encompassed 8 counties was divided into two four-county districts. This reflects the continued settlement of these last settled areas of Ohio. Even districts comprised of four counties are larger than the normative two or three counties.

The Toledo district was also divided. This is largely based on the growth of Toledo, likely due to it being the closest lake port to northwestern Ohio and northeastern Indiana, and even portions of southeastern Michigan.

One district, Lorain-Medina fell below the 3.0 threshold and the two counties were assigned to adjacent districts.

Hamilton County would have three districts, two wholly in Cincinnati and one partially in the city. In 1860, it was possible to create a district outside Cincinnati, as Mill Creek township just north of the original settlement along the river had over 10K population. Much of it was annexed into the city by 1870. In modern times only a few enclaves such as Norwood have not become Cincinnati. By 1870, Springfield and Sycamore townships exceeded 5K. This might be due to growth from the city, or perhaps higher density commercial farming to feed the city. In 1870, Cincinnati would be entitled to 11.2 representatives and the remainder of the county 2.4.

Cuyahoga would have two districts, with Cleveland entitled to 5.0 representatives and the remainder of the county 2.0. Notable areas outside the city include East Cleveland and Newburgh townships. East Cleveland township is now in the cities of East Cleveland, Cleveland Heights, parts of Euclid, and Bratenahl, but mostly in eastern Cleveland. Newburgh was southeast of the city center along the east bank of the Cuyahoga River.


Largest cities in 1870

Cincinnati 216K (8th largest in USA)
Cleveland 93K (15th)
Toledo 32K (40th)
Columbus 31K (42nd)
Dayton 30K (44th)
Sandusky 13K
Springfield 13K
Hamilton 11K
Portsmouth 11K
Akron 10K
Zanesville 10K
Delaware 10K
Chillicothe 9K
Canton 9K
Steubenville 8K
Youngstown 8K
Mansfield 8K
Newark 7K
Xenia 6K

Toledo had a explosive growth as it developed into a mini-Chicago with rail lines radiating out from a lake port. Akron, Youngstown, and Canton also appeared on the list as industrialization begins to take hold (due to war production?).

Ohio's first railroads were around 1850, but had developed into a robust network by 1870 permitting growth of inland towns.

Railroads from the east were the water level route from New York City up the Hudson and then west along the route of the Erie Canal to Buffalo and down along Lake Erie to Cleveland; The Pennsylvania Main Line from Philadelphia to Pittsburgh, and the Baltimore & Ohio, up the Potomac River crossing into Ohio at Wheeling and Parkersburg (one reason for West Virginia was to keep the B&O in Union control, and in particular accounts for the eastern panhandle where the route is on the south side of the river).

Once the Alleghenies had been conquered the terrain was more forgiving, and a complex network of railroads was created as the three narrow routes from the east branched out unwilling to concede cities such as Cincinnati and Chicago, etc. to competitors.

A counter example of a failed city is Pomeroy which has 1/4 of its 1860 population (6.8K to 1.7K) which is in Meigs County on the upper bend in the Ohio River just to the west of that comma shaped peninsula. I did find a reference to coal mining, though Meigs is east of the Ohio coal belt. Perhaps there were shallow beds which were easier to mine, but were not of the quality to compete with Eastern Kentucky coal. That is what happened to the coal-mining areas of Iowa, which were abandoned once rail lines were built to the Kentucky mines.

It may have also suffered from repeated flooding. This spring, flooding was expected to reach 48 feet, which was lapping at the doors along Main Street. But the historic crests have been as high as 68 feet, which would put two-story buildings under water and the first floor for perhaps weeks. That will ruin wood floors and perhaps collapse brick buildings.


372K foreign born (14.0%)
450K born outside Ohio (16.9%)
Ohio-born 69.9%

Of the native born of foreign-parentage:

359K foreign mother and foreign father (75.2%)
85K foreign father and native mother (17.9%)
33K foreign mother and native father (6.9%)

It is interesting that children born to a foreign father and native mother are much more common than the reverse.

Possible reason?

Surplus of males among immigrants? Second marriages by a widow/widower more likely to produce additional children if the father is 40 than if the mother is 40? Greater social acceptability of older groom, younger bride marriages than the reverse? An expectation that a father be economically established before marriage?

Most adult immigrants are likely to be late teens or twenties. Older adults with a large number of children may be more settled and less likely to emigrate even if they are in dire circumstances. They might send a teenager off to live with extended family in America.

But this would lead to marriages being formed in Europe or arranged in Europe or a marriage within a few years of immigration when one is most dependent on their cultural group, so that the children are born to two foreign parents. Some of the native born with two foreign parents may be 50 years old and have grandchildren of their own. Others, in 1870 may be too young to have children of their own, even if they are still affiliated with their ethnic community.

If we assume that all foreign born had two foreign-born parents, then there are 477K natives of with at least one foreign parent. This is 17.9% of the total. Combined with the foreign born they comprise 34.8% of the Ohio population is 1st or 2nd generation (foreign stock).

Among the foreign-born

49.1% German states.
22.2% Ireland.
9.8% England.
3.5% Canada/British North America
3.5% Wales
3.4% Switzerland
3.4% France

At this date (1870) immigrants from Italy and Slovakia were negligible.

In 1870 of total population:

Cincinnati: 22.9% German, 8.6% Ireland
Cleveland: 17.1% German, 10.7% Ireland, 2.8% (Canada)
Toledo: 16.9% German, 12.4% Ireland, 3.2% Canada
Columbus: 12.7% German, 5.9% Ireland
Dayton: 16,3% German, 4.4% Ireland

Nothing too surprising, people migrate to work, and there are more jobs in the growing cities. Columbus and Dayton may be attracting Ohioans migrating in from the farms. If you want to be a farmer you migrate across the Mississippi.


Among persons born in other States:

PA: 33.3%
NY: 15.0%
KY: 5.8%
MD: 5.2%
VA/WV: 4.2%
IN: 3.9%

Compared to 1860, IN is new to the list. There may have been enough westward migration to develop significant migration back to home. At least in modern times this is fairly common.

KY and VA/WV migration had a significant black component (around 1/3). There was negligible migration from more southerly states, white or black. The migration from Maryland was white, likely from the wheat growing areas of northern Maryland. Former slaves likely headed north to Philadelphia or New York.

Cuyahoga was overwhelmingly from NY (around 40%), while Hamilton was KY, PA, and NY, with KY having a slight lead.

Is it simple to distinguish people between people from Cincinnati and Cleveland? Are people from Cleveland rude while those from Cincinnati are more genteel?


The 1870 distinguishes among German states, it is interesting how much differentiation there is:

Among the 50 largest cities:

Austria (proper): NYC, Cleveland. Since Bohemia and Hungary are separately listed, would this include or not include Slovaks?
Bohemia: Chicago
Canada: Chicago, Detroit (this might be only Ontario)
Maritimes (NB, NS, PE were listed separately, along with Newfoundland): Boston
Remainder of British North America (this may be Quebec): Boston
Denmark: Chicago
France: New Orleans, NYC

Baden: Philadelphia
Bavaria: NYC, Cincinnati, Baltimore
Hanover: St. Louis, Cincinnati
Hessen: NYC, Baltimore
Oldenburg: Cincinnati
Prussia: New York, St. Louis Chicago
Wurtemburg: Philadelphia

England: NYC, Philadelphia, Brooklyn
Ireland: NYC, Philadelphia, Brooklyn, Boston
Wales: Scranton
Italy: NYC, San Francisco, New Orleans
Norway: Chicago
Poland: NYC, Chicago
Portugal: Boston
Russia: NYC
Sweden: Chicago
Switzerland: St. Louis, NYC, Philadelphia


50 largest cities distributed among States:

8 NY
8 MA
5 PA
5 OH
3 NJ
2 MO
2 CT
17 with 1: IL, MD, LA, CA, DC, KY, MI, WI, RI, VA, SC, IN, TN, AL, ME, DE, GA

The 17 are listed in rank order of their largest city. You should be able to guess all of these except perhaps GA.

Three of the top 50 no longer exist as a separate city. Cities in 6 of the 24 states is no longer even the largest city in their State.

It took 26,766 to qualify for the Top 50 cities in 1870.

[jimrtex]

In the 1870's there are 12 floating representatives (of a total of 139). Districts are grouped in geographical areas. One representative position floats in each of eight areas. Two representatives float in each of two areas. In addition one representative may float among the three districts in Hamilton County and two districts in Cuyahoga.




As in the 1860's the extra positions are assigned in a way that is randomized with the goal of preventing a district from having (or not having) extra representatives in consecutive sessions, to the extent possible, and to also avoid districts with the same number of extra sessions having those extra sessions in sync.

1872	1874	1876	1878	1880
Darke	Butler	Darke	Darke	Darke
Ross	Ross	Ross	Ross	Franklin
Seneca	Lucas	Lucas	Lucas	Lucas
Erie	Erie	Seneca	Erie	Seneca
Columbiana	Stark	Columbiana	Stark	Columbiana
Stark	Trumbull	Trumbull	Ashtabula	Trumbull
Washington	Meigs	Washington	Meigs	Meigs
Knox	Muskingum	Licking	Muskingum	Muskingum
Hamilton	Clermont	Hamilton	Hamilton	Clermont
Lawrence	Highland	Lawrence	Lawrence	Highland
Allen	Williams	Williams	Allen	Williams
Miami	Champaign	Montgomery	Champaign	Montgomery

This is the presentation used for the general public. At a glance they can find their district, and the number of representatives elected in each session. There would be insets showing the three districts in Hamilton County and two districts in Cuyahoga.

[jimrtex]

This is the 1880 Apportionment for Ohio.



Growth during the decade resulted in the expansion of the legislature according to the cube root rule from 139 to 147 members. Statewide growth was 20.0%. To maintain its level of representation a district would need to increase by 13.5%.

The statewide growth during the 1870s increased from 13.5% during the the 1860s. This might have been due to premature deaths during the Civil War, or deferred births due to potential fathers being away, or burgeoning industrialization providing jobs in cities large and small. Up to 80 years into settlement, farming areas had likely reached their capacity. In the past, excess population might have moved further west. Now there would be more jobs available.


Lucas (Toledo) had surpassed the 3.0 threshold and became its own district. Wood was detached and added to the Erie district, with Wood becoming the most populous county in the district.

Clark (Springfield) became the largest county in its district surpassing Miami. Springfield was now the sixth largest city in Ohio (and 94th largest in the US).

No districts fell below the 3.0 threshold, though nine districts were at 3.2 or 3.0.

Stark (Canton and Massillon) was close to reaching the 3.0 threshold in order to become its own district. This will likely happen in 1890.

Hamilton would be required to have three districts, all would have required part of the city of Cincinnati. Cincinnati would be entitled to 11.8 representatives, while the portion of the county outside the city would be entitled to 2.6 representatives. One formulation might be to place an area equivalent to 0.8 representatives with the county area for a total of 3.4 representatives, while the remaining 11.0 area in the city would be divided into two districts. This would permit the representative from the county area to float with other counties, while another representative would be permitted to float between the two Cincinnati districts.

Cuyahoga would be required to have two districts, both of which would have required a portion of Cleveland. Cleveland would be entitled to 7.4 representatives would be entitled to 1.6 representatives. A division might be made at the Cuyahoga river, with the county area divided between the two districts.


Largest cities in 1880:

Cincinnati 255K (8th largest in USA)
Cleveland 160K (11)
Columbus 52K (33)
Toledo 50K (35)
Dayton 39K (47)
Springfield 21K (94)
Zanesville 18K
Akron 16K
Sandusky 16K
Youngstown 15K
Steubenville 13K
Canton 12K
Hamilton 12K
Portsmouth 11K
Chillicothe 11K

There are 22 other cities that exceed 5K. 36 of 88 counties (41%) have a city of over 5K which would be the commercial center (and usually the government center) for the county. A typical township would have around 1K (if a township were 6 miles by 6 miles that corresponds to a density of 23 acres/person).

Summit (Akron) has not yet begun to differentiate in a way that would make it one of six large Ohio cities (Cincinnati, Cleveland, Columbus, Dayton, Toledo, and Akron). Charles Goodyear had discovered the vulcanization process that hardens rubber in 1839, but was not financially successful.

The Goodyear Tire and Rubber Company was founded in Akron in 1898, and named in honor of Charles Goodyear. The Seiberling State Tire Building machine patented in 1908 increased production of tires per 10 man-hours from 5 to 60 and in conjunction with mass auto production pushed growth of Akron.


395K foreign-born (12.3%)
442K born outside Ohio (13.8%)
Ohio-born 73.8%

This was an almost 5% increase in Ohio-born. The share of foreign-born and non-Ohio natives decreased. There was a slight numeric increase in the number of foreign-born.

These trends likely reflect the aging of the agricultural population. The original settlers likely came from out of the state. Many of these farms may have passed to their Ohio-born children. These children and grandchildren may have also moved into the cities. The excess farm population from the east would now be moving past the Mississippi, skipping over Ohio.

The foreign-born population was substantially higher in Cleveland, Cincinnati, and Toledo as these rapidly growing cities provided jobs in manufacturing. Were it not so close to Detroit, Toledo might have grown to rival Cincinnati and Cleveland.

Cincinnati: Ohio 59.4%, Native 12.5%, Foreign 28.1%
Cleveland: Ohio: 49.0%, Native 13.3%, Foreign 37.1%
Columbus: Ohio: 67.5%, Native 14.9%, Foreign 17.8%
Toledo: Ohio 52.4%, Native 19.0%, Foreign 28.6%
Dayton: Ohio 65.8%, Native 15.5%, Foreign 18.7%

Leading birthplace on non-Ohio: PA, NY, VA, KY, and IN.
Leading destination of Ohio-born: IN, IL, IA, KS, and MO.

Ohio was the #1 destination for those born in PA and WV.
It was #3 for those born in MD, MI, and VA
It was #4 for those born in CT, NJ, and DE
It was #5 for those born in IN

Sources for

Cincinnati: KY, IN, NY, PA
Cleveland: NY, PA, MA, MI
Columbus: PA, NY, VA, KY
Toledo: NY, MI, PA, IN
Dayton: PA, NY, IN, KY


Among persons born in other states:

PA: 42.6%
NY: 14.5%
VA: 11.7%
KY: 7.4%
IN: 6.2%
MD: 4.5

Among blacks:

OH: 55.8%
VA: 15.2%
KY: 14.5%
NC: 3.1%
TN: 2.0%
PA: 1.4%
WV: 1.4%

It is possible that some births in WV were reported as being in VA, since that might have been true at the time of birth. Enumerators may not have made close inquiries in county of birth - and some may not have known.


Sources of foreign born:

Germany 48.8%
   Unspecified: 13.7%
   Prussia: 10.4%
   Bavaria: 6.2%
   Baden: 5.2%
   Hanover: 3.9%
   Wurttemburg: 3.9%
   Hessen: 2.8%
Ireland 20.0%
England 10.5%
British North America: 4.1%
Wales 3.5%
Switzerland 3.0%
Scotland 2.3%
Bohemia 1.6%

Close to 95% of the Bohemian-born were living in Cuyahoga (Cleveland).

Large communities of foreign born:

Germany: Hamilton 53K, Cuyahoga 27K, Lucas 8K, Montgomery 8K, Franklin 6K
Ireland: Hamilton 17K, Cuyahoga 13K
England: Cuyahoga 11K
Bohemia: Cuyahoga 6K

While Toledo, Columbus, and Dayton may have significant Irish population share, the cities were not large enough to have enough population to support community institutions (5K as a threshold is fairly arbitrary)


50 Largest cities.

7 NY (-Utica)
7 MA (-Charletown, annexed to Boston)
5 PA
5 OH
4 NJ (+Camden)
2 MO
2 CT
2 MN (+Minneapolis, +St.Paul)
1 IL
1 MD
1 CA
1 LA
1 DC
1 KY
1 MI
1 WI
1 RI
1 IN
1 VA
1 SC
1 TN (+Nashville, -Memphis)
1 DE
1 GA (+Atlanta, -Savannah)
1 CO (+Denver)

[jimrtex]

In the 1880's there are 11 floating representatives (of a total of 147). Districts are grouped in geographical areas. One representative position floats in each of nine areas. Two representatives float in one area in the central area (one session has both in the western subarea: Champaign and Clark, skipping the eastern subarea of Licking and Muskingum. In addition one representative may float among the three districts in Hamilton County and two districts in Cuyahoga.



This is the constrained random assignment of the extra sessions for the 11 floating seats.

1882	1884	1886	1888	1890
Seneca	Seneca	Lucas	Seneca	Seneca
Columbiana	Columbiana	Columbiana	Columbiana	Ashtabula
Lawrence	Clermont	Lawrence	Lawrence	Lawrence
Stark	Stark	Stark	Summit	Stark
Darke	Greene	Greene	Greene	Greene
Clark	Champaign	Clark	Champaign	Champaign
Licking	Muskingum	Licking	Muskingum	Clark
Montgomery	Hamilton	Montgomery	Montgomery	Hamilton
Highland	Ross	Ross	Highland	Ross
Meigs	Meigs	Belmont	Meigs	Washington
Allen	Wood	Hancock	Wood	Hancock

This is the number of apportioned representatives for each session.

[jimrtex]

This is the 1890 Apportionment for Ohio.



Growth during the decade resulted in the expansion of the legislature according to the cube root rule from 147 to 154 members. Statewide growth was 14.8%. To maintain its level of representation a district would need to increase by 9.6%.

The statewide growth during the 1880s (14.8%), declined from 20.0% in the previous decade reflecting a maturing of the population and little immigration into agricultural areas. While the cities were growing, much of this population was from the countryside. Most other states had their own urban centers, and Ohio may have been losing residents to Detroit and Chicago.

Stark (Canton and Massillon) had surpassed the 3.0 threshold, and Carroll was detached.

Clermont, Knox, and Washington districts fell below the 3.0 threshold. The districts were dissolved and the constituent counties assigned to adjacent districts.

Clermont and Brown were the largest counties in their new districts, replacing Greene and Highland, respectively.

Defiance, Logan, Athens, and Mahoning became the largest county in their districts, replacing Williams, Champaign, Meigs, and Trumbull, respectively.

It would be possible to draw a Hamilton County district outside Cincinnati with a population equivalent to 3.2 representatives. Cincinnati would be entitled to 12.6 representatives divided among three districts.

Cuyahoga would be entitled to three districts. The area outside Cleveland would only be entitled to 2.0 representatives, it is possible that areas outside the city would be assigned to all three districts, particularly since the Cuyahoga River provides a natural boundary.

In 1850, there was only one single-county district, while most districts were two counties. By 1890, there were six single-county districts, while four counties had become normative for less urban areas.

[jimrtex]

In the 1890's there are 10 floating representatives (of a total of 154). Districts are grouped in geographical areas. One representative position floats in each of ten areas. In addition a few representatives may float among the four districts in Hamilton County and three districts in Cuyahoga.




This is the constrained random assignment of the extra sessions for the 10 floating seats.

1882	1884	1886	1888	1890
Seneca	Seneca	Lucas	Seneca	Seneca
Columbiana	Columbiana	Columbiana	Columbiana	Ashtabula
Lawrence	Clermont	Lawrence	Lawrence	Lawrence
Stark	Stark	Stark	Summit	Stark
Darke	Greene	Greene	Greene	Greene
Clark	Champaign	Clark	Champaign	Champaign
Licking	Muskingum	Licking	Muskingum	Clark
Montgomery	Hamilton	Montgomery	Montgomery	Hamilton
Highland	Ross	Ross	Highland	Ross
Meigs	Meigs	Belmont	Meigs	Washington
Allen	Wood	Hancock	Wood	Hancock

This is the number of apportioned representatives for each session.