City of Hudson's weighed voting system under scrutiny

[Kevinstat]

I was thinking about your desire to maintain the status quo. The status quo is that the residents of Westwind vote in Ward 4. So how about this alternative?

I like that map, particularly from a selfish political standpoint (well other than for me), but again, we are not crossing Warren Street, and I want to minimize population variances.

How about this?



I used Jimrtex's map above as a basis, with my modifications in Microsoft Paint.  You can ignore where the black lines are.  The background colors are what rules.  Wards 1 and 2 are exactly as in the current Torie plan, and the populations of the other three wards are the same, just switched around.

[jimrtex]

I was thinking about your desire to maintain the status quo. The status quo is that the residents of Westwind vote in Ward 4. So how about this alternative?

I like that map, particularly from a selfish political standpoint (well other than for me), but again, we are not crossing Warren Street, and I want to minimize population variances.

How about this?



I used Jimrtex's map above as a basis, with my modifications in Microsoft Paint.  You can ignore where the black lines are.  The background colors are what rules.  Wards 1 and 2 are exactly as in the current Torie plan, and the populations of the other three wards are the same, just switched around.

My problem is that it divides (cracks) an apartment complex with a large minority population, and utilizes an alleyway as a ward boundary.

[jimrtex]

Added 6th Ward

Hudson has always divided wards, rather than modifying ward boundaries. In 1815, Hudson was divided on 3rd Street into two wards. In 1855 these wards were divided along Warren Street (and Columbia Street and Columbia Turnpike) creating four wards. In 1885, Ward 4 was divided on N 5th Street extended. Since Ward 5 now has a dominant share of the population, it could be divided on State Street and the railroad tracks. The remainder of the 5th Ward and the new 6th Ward would each elect two aldermen, while Wards 1 and 4 would election one each.

Interactive Version

[jimrtex]

This would be a better starting point for drawing maps.



It begins with my suggested blocks, and then fuses blocks divided by alleys, with two exceptions: Strawberry Alley, which forms a substitute for an escarpment; and Deer Alley which has facing residences and functions as a street.

It also fuses blocks containing apartment complexes (Hudson Terrace and Crosswind).

It also fuses some uninhabited and uninhabitable areas around the edge of the city.

Conceivably, separate areas could be devised for areas where habitation is distinct. For example, the brown northern block could have its Mill Street, Lucille Drive and Firemen's Home population separated.

[Kevinstat]

I was thinking about your desire to maintain the status quo. The status quo is that the residents of Westwind vote in Ward 4. So how about this alternative?

I like that map, particularly from a selfish political standpoint (well other than for me), but again, we are not crossing Warren Street, and I want to minimize population variances.

How about this?



I used Jimrtex's map above as a basis, with my modifications in Microsoft Paint.  You can ignore where the black lines are.  The background colors are what rules.  Wards 1 and 2 are exactly as in the current Torie plan, and the populations of the other three wards are the same, just switched around.

My problem is that it divides (cracks) an apartment complex with a large minority population, and utilizes an alleyway as a ward boundary.

While I welcome your input, my posting of this map was more aimed at Torie.  It gets rid of the split across at least one home in the block around Underhill Pond, while still only shifting territory 4 ways (3 -> 1, 5 -> 3, 5 -> 4 and 2 -> 4).  I wanted to find something you would object to less than the map Torie is currently advocating for, while meeting Torie approval which I placed a higher value on since he's the one who lives in Hudson and is involved "on the ground" in the petition drive to get a five equal population wards (and presumably ditching weighted voting) scheme adopted.

It's probably too late now for Torie to be willing to consider changes in his map based on where he is in the process, but I thought I'd throw this out there to see what he thought.  Who knows, he might be getting pushback about keeping the extension of N 5th St as a ward boundary between Clinton St and Harry Howard Ave, or "Why did you keep that extension [part of the way] but not the 3rd Street extension?", and my alternative could help address those issues.

[jimrtex]

While I welcome your input, my posting of this map was more aimed at Torie.  It gets rid of the split across at least one home in the block around Underhill Pond, while still only shifting territory 4 ways (3 -> 1, 5 -> 3, 5 -> 4 and 2 -> 4).  I wanted to find something you would object to less than the map Torie is currently advocating for, while meeting Torie approval which I placed a higher value on since he's the one who lives in Hudson and is involved "on the ground" in the petition drive to get a five equal population wards (and presumably ditching weighted voting) scheme adopted.

It's probably too late now for Torie to be willing to consider changes in his map based on where he is in the process, but I thought I'd throw this out there to see what he thought.  Who knows, he might be getting pushback about keeping the extension of N 5th St as a ward boundary between Clinton St and Harry Howard Ave, or "Why did you keep that extension [part of the way] but not the 3rd Street extension?", and my alternative could help address those issues.

At this point it would probably have to be the Common Council that would propose an alternative.

The group that is pushing this is from a small area, and is likely promoting a gerrymander under the  guise of "equality". Small councils are the easiest to gerrymander.

[jimrtex]

10 Wards

With 10 wards, each would elect one alderman.  This will provide a more geographically diverse council, and likely more politically diverse, since each alderman will be elected by a unique electorate. The existing wards are retained, except Harry Howard Avenue is substituted for the confusing 5th Street extension and Columbia Street is substituted for Columbia Turnpike, conforming to current practice.

The wards could be named. Perhaps an election among residents over a certain age (say 8 YO), could choose the name of their ward.

Interactive Version

[jimrtex]

Alternative 1

The current proposal has three significant problems:

(1) Cracking of the Hudson Terrace apartments, which has a large minority population;
(2) Maintaining the 5th Street extension as a ward boundary, which has caused so much confusion in the past;
(3) Using alleys as ward boundaries, which suggests deliberate gerrymandering, and is likely to cause confusion.

This alternative places all of Hudson Terrace in Ward 2; uses Harry Howard Avenue as the boundary between wards 4 and 5; and eliminate use of alleys.

Wards are numbered to recognize that the ward south of Warren Street is a successor to both wards 1 and 3, with Ward 3 the larger contributor; and that 70% of then new Ward 6 is from Ward 5.

[jimrtex]

Alternative 2

This alternative cleans up the area east of 5th Street.

[jimrtex]

Alternative 3

This version shifts the area north of Underhill Pond into Ward 4.

[jimrtex]

Alternative 5

This retains areas east of Worth Avenue in Ward 3.

[jimrtex]

Alternative 6

The moves the Firemen's Home into Ward 5, which simplifies the area east of 5th Street.

[jimrtex]

Alternative 7

This is based on the current wards. Ward 5 is currently just short of enough population for two wards, while Ward 3 is about 200 short of the quota. Rather than swapping people around and chopping up areas, this plan preserves the wards which have sufficient population.

Ward 3 is extended westward to 2nd Street, while Ward 5 picks up enough population by switching to Harry Howard Avenue as its western boundary. Ward 5 is then divided into Wards 5 and 6.

Because Ward 4 is short, it must also shift to the west, rather than piecemeal additions to the east and west. Ward 2 then takes the remainder of Ward 1.

The plan has the smallest deviation, and the fewest persons moved between wards. It doesn't use artificial boundaries, crack Hudson Terrace, or use alleys as boundaries.

[jimrtex]

Six Equal Population Districts

For its size, Hudson has a large city council, and multi-member districts tend to reduce diversity. Hudson could increase the number of wards, and make them single member.

With six wards electing one alderman, the council president would be the 7th member of the common council.

Or perhaps there could be three aldermen elected at large, using some pluralistic system such as cumulative voting. A simplified version would let a voter vote for one, two, or three candidates. If they voted for three, each candidate would get two votes; if they voted for two, each would get three votes; if they voted for one, he or she would get six votes. The three top vote-getters would be elected.

Thus a voter would have both geographical representation, and political representation that would not necessarily be geographically based.

[jimrtex]

Seven Equal Population Wards

With seven wards, it might be better to eliminate the council president.

[jimrtex]

Eight Equal Population Wards

With 8 wards, the council president would make for a 9th member.

[jimrtex]

BEST PLAN



VS

TORIEMANDER



Equality (Relative Deviation Range):

Best Plan 1.41% Toriemander 1.87%

Equality (Standard Deviation):

Best Plan 0.60% Toriemander 0.78%

Retained Boundaries:

Best Plan 2.11 miles Toriemander 1.54 miles.

The Best Plan retains part of the 1815 3rd Street boundary (Harry Howard was also a boundary at the time); most of the 1855 North/South boundary; and most of the actual 5th Street portion of the 1885 boundary.

Shifted Persons:

Best Plan 1256 (19.6%) Toriemander 1457 (22.8%)

Split Census Blocks:

Best Plan 1  Toriemander 2

The best plan uses a railroad track that was a block boundary in 2000, and the census bureau plans to use in 2020. The Toriemander splits Hudson Terrace apartments, a residence, and Underhill Pond.

Alleyways

Best Plan 0  Toriemander 2

The best plan uses Deer Alley which has houses fronting it.

[jimrtex]

You must be logged in to read this quote.

Did this happen?

[jimrtex]

These are interesting cases, in the Supreme Court (ie a New York district court)

Matter of Curcio v Boyle 142 Misc.2d 1030 (1989)

and on appeal.

Matter of Curcio v Boyle 147 A.D.2d 194 (1989)

Citizens in Suffolk County had initiated a proposal to switch from a legislature comprised of single-member districts to a board of supervisors using weighted voting. The Suffolk County Charter provides for an initiative procedure. The County Attorney (Boyle) initially approved the petition as sufficient. He later he reversed himself and made a number of claims. The Supreme Court judge wryly observed, "[t]here is no indication as to how these tidbits previously escaped the attention of the County Attorney."

However, the appeals court ultimately rejected the initiative because it did not provide enough details of how the voting weights would be calculated.

"the modified weighted voting standards enunciated and approved by the Court of Appeals in Iannucci v Board of Supervisors, 20 N.Y.2d 244 * * * and in Franklin v Krause, 32 N.Y.2d 234" does not adequately apprise the voters from any of the towns within Suffolk County of precisely how many votes each of their respective Supervisors would cast."

If the petition that the Hudson voters signed did not include the full text of the proposed local law, is the petition legally sufficient?

Paragraph 2 of MHR Section 37 makes clear that it is the signers of the petition who are filing the petition with city clerk. A couple of individuals carrying the signed sheets into the clerk's office are merely agents of the signers. Paragraph 2 requires the proposed law to be "set forth in full" in the petition.

Paragraph 3 requires the additions to be in italics or underlined and deletions to be indicated with strike-throughs or [brackets]. In essence, an initiative is a legislative act, where those who are supporting placing the matter before the governing body the voters are fully informed as what they are supporting.

If the circulated text which voters signed included the entire petition, as required by law, I don't understand why Torie would not let me see it while the petitions were being circulated.

[Kevinstat]

These are interesting cases, in the Supreme Court (ie a New York district court)

Matter of Curcio v Boyle 142 Misc.2d 1030 (1989)

and on appeal.

Matter of Curcio v Boyle 147 A.D.2d 194 (1989)

Citizens in Suffolk County had initiated a proposal to switch from a legislature comprised of single-member districts to a board of supervisors using weighted voting. The Suffolk County Charter provides for an initiative procedure. The County Attorney (Boyle) initially approved the petition as sufficient. He later he reversed himself and made a number of claims. The Supreme Court judge wryly observed, "[t]here is no indication as to how these tidbits previously escaped the attention of the County Attorney."

However, the appeals court ultimately rejected the initiative because it did not provide enough details of how the voting weights would be calculated.

"the modified weighted voting standards enunciated and approved by the Court of Appeals in Iannucci v Board of Supervisors, 20 N.Y.2d 244 * * * and in Franklin v Krause, 32 N.Y.2d 234" does not adequately apprise the voters from any of the towns within Suffolk County of precisely how many votes each of their respective Supervisors would cast."

If the petition that the Hudson voters signed did not include the full text of the proposed local law, is the petition legally sufficient?

Paragraph 2 of MHR Section 37 makes clear that it is the signers of the petition who are filing the petition with city clerk. A couple of individuals carrying the signed sheets into the clerk's office are merely agents of the signers. Paragraph 2 requires the proposed law to be "set forth in full" in the petition.

Paragraph 3 requires the additions to be in italics or underlined and deletions to be indicated with strike-throughs or [brackets]. In essence, an initiative is a legislative act, where those who are supporting placing the matter before the governing body the voters are fully informed as what they are supporting.

If the circulated text which voters signed included the entire petition, as required by law, I don't understand why Torie would not let me see it while the petitions were being circulated.

While I'm not thrilled about Torie's map, I'd like to see the referendum happen, so if you can avoid trying to tip off the opposition so they can squelch the referendum, I'd appreciate it.

[jimrtex]

Torie's group made a presentation to the Common Council regarding their initiative. Other business takes up a few minutes at the beginning of the meeting.

Hudson Common Council 11 July 2016

The primary concern was that two black supervisors were paired under the plan, including one who has served for 20 years.

The two aldermen from the 2nd Ward were generally ranting, but it was difficult to understand Alderman Miah because of his accent. Alderman Garriga expressed concern about Robinson and Mill Street being moved into the 4th Ward, but her concern may have really been Robinson Street.

I suspect that if you set either down with a map with block populations that neither would be willing (or able) to identify which changes could be made to Ward 2, to bring it into within equal population range.

Supervisors Cross and Hughes at least offered cogent alternatives, though Hughes made it sound like the voting by Crosswinds was deliberate, and that there would therefore be legal culpability.

Incidentally, Cross may have misunderstood the 148 persons on Robinson to not include both sides of the street. Four persons were added by LATFOR, so the Census Bureau counted 144. The number of housing units reported by the Census Bureau matches property records reasonably. The north side has 3 housing units more than the census, and the south side has one less. But I think a "housing unit" does not have to be fully legal. The population per unit in the census is reasonable, with larger minority households, which would be consistent with families with children being more likely to be minority.

Why couldn't Hudson conduct a census for the split blocks?

You are right about Alderman Friedman. Kudos for the Council President in running the meeting with equanimity.

BTW, the whole situation shows why equal population districts are evil. I suspect that Justice Brennan when analyzing the Judgment of Solomon, said, "I understood it until that emotional lady intervened and disrupted the proceedings."

Torie is wrong about weighted voting being unconstitutional.

You must be logged in to read this quote.

Doesn't this preclude the City of Hudson from changing the districts from which supervisors are elected? What, if any, duties do supervisors perform in their capacity as city officers? Do they subscribe to two oaths of office?

Why do the supervisor districts have to be changed? The supervisors execute no legislative duties in Hudson.

Would this satisfy supervisors Hughes and Cross.



Torie has not actually presented a reason for not crossing Warren. The block between 1st and 2nd Street, Warren Street and Columbia Street has the smallest minority proportion, and placing it in Ward 1/3 avoids cracking Hudson Terrace apartments.

The Mill Street area could conceivably be placed in Ward 2. There is a justification in that there is physical way to get to the area by automobile from the remainder of the 4th Ward. The deviation for Ward 2 would be +5.1% and overall range of 8.9%. But the VRA angle is problematic.

The placement of the Firemen's Home in the 5th Ward is indicative. A legal description would either use the address of the Firemen's Home, or a description "all properties on the west side of Harry Howard Avenue, north of Paddock Place", or use the property boundary of the Firemen's Home (and the High School).

[Kevinstat]

When Hudson was split between two wards along 3rd Street in 1815, what is now the town of Greenport was part of Hudson, according to Wikipedia.  How did the boundary extend north and south of what is now Hudson?  S 3rd Street (at least now) curves (as State Routes 23B and 9G) when you get into Greenport, and the extension of N 3rd Street into Lorenz Park and Stottville does cross some present roads (including what is now U.S. 9 in the Greenport section of Stottville).  Perhaps back then there were no homes in then-Hudson north of present-day Hudson on the northeast side of the 3rd Street extension.  But in the south, well I'd need to see a map from around 1815 to see what the logical extension of South 3rd Street.  Was there a bridge over the Hudson River in (then) southern Hudson and Catskill back then?  The road (and the 1815 ward boundary before Greenport left in 1837) could have also followed (at least roughly) the course of the present 9G into Livingston, even if there was a forking as there is now between a road crossing the Hudson and one continuing roughly parallel to the river (there's a forking going the other direction to, but no one could argue that the present-day State Route 23 going east and then southeast would be any kind of extension of 3rd Street.

[jimrtex]

When Hudson was split between two wards along 3rd Street in 1815, what is now the town of Greenport was part of Hudson, according to Wikipedia.  How did the boundary extend north and south of what is now Hudson?  S 3rd Street (at least now) curves (as State Routes 23B and 9G) when you get into Greenport, and the extension of N 3rd Street into Lorenz Park and Stottville does cross some present roads (including what is now U.S. 9 in the Greenport section of Stottville).  Perhaps back then there were no homes in then-Hudson north of present-day Hudson on the northeast side of the 3rd Street extension.  But in the south, well I'd need to see a map from around 1815 to see what the logical extension of South 3rd Street.  Was there a bridge over the Hudson River in (then) southern Hudson and Catskill back then?  The road (and the 1815 ward boundary before Greenport left in 1837) could have also followed (at least roughly) the course of the present 9G into Livingston, even if there was a forking as there is now between a road crossing the Hudson and one continuing roughly parallel to the river (there's a forking going the other direction to, but no one could argue that the present-day State Route 23 going east and then southeast would be any kind of extension of 3rd Street.

At the time, Hudson included part of Stockport as well. The original northern boundary was Stockport Creek (originally known as Major Abram's Creek) which is a short east-west creek from the junction of Kinderhook Creek and Claverack Creek. Kinderhook Creek flows southwest from Kinderhook and has a pretty deep gorge. Claverack flows northward and was the original eastern boundary of Hudson, and is now the eastern boundary of Greenport. Though parallel to the Hudson River, they flow in opposite directions as Claverack Creek is seeking a way to drain into the Hudson River. The southern boundary of Hudson was Livingston, Livingston was originally Livingston Manor, granted by Queen Anne.

I'm pretty sure the crossing of Claverack Creek was in (present Stockport). Based on the description of the boundaries of Ghent, the crossing was at Stottville. So the ward boundary would have barely got into modern Stockport and ran eastward.

The Dugway is from just off the northern tip of N 3rd Street up to what is Harry Howard Avenue. Harry Howard was a legendary head of the volunteer fire department in New York City. The road in Hudson was named for him after the Firemen's Home was built. Harry Howard runs into Joslen Boulevard just outside the city limits, and I'm pretty sure this would have been the ward boundary. The straighter route of US 9 was the Farmer's Turnpike. Just into Greenport there is an old Dutch house off to the west, which dates from the 17th Century. Before there was a Hudson, there was a Claverack Landing at both North Bay and South Bay. If you were coming from the north, it would be easier (not easy) to go down the Dugway to the North Bay. Only later was the road built below Underhill Pond.

The Highland Turnpike was somewhat of a boondoggle, or perhaps simply an ambitious project that didn't quite pan out. It began at Kingsbridge, which spanned the Harlem River between Manhattan and the Bronx. It was intended to reach Albany (or maybe Troy). There were a lot of cost-overruns such as building a bridge at Croton-on-Hudson (over a Hudson tributary), and it ran into competition from steamboats on that north-south route. In Hudson the turnpikes to the east, Columbia and Union were more successful because they only had competition from public roads and agricultural goods could be fed into Hudson for consumption there, or shipped downriver to NYC. I think I've seen a petition from the citizens of Hudson to revoke the charter. Anyhow, it was never really established in Hudson.

I think I figured out a path more southeasterly once it got out of the city, but I was working off guidebooks for suggested bike tours. Around 1900 the "Good Roads" advocates were bike enthusiasts.

Blue Store to Hudson via Blue Hill Road (Google Map).

The bridges over the Hudson are pretty new. The Rip Van Winkle Bridge was built in 1935. Many of the bridges are likely Public Work era. They have about 150 foot clearances, and some were superlative for their time (longest suspension bridge, highest, etc.). If you wanted to cross the river you would use a ferry. The bill that created the wards, also permitted Hudson and the village of Athens to operate a ferry, and dig a canal through the Middle Ground (the island between the burgs). At low tide, the ferry had to take much more southerly looping route.

I don't think the ward boundary was changed when Stockport and the Greenport were detached. So the ward boundary would have gone up the Dugway and out (now) Harry Howard Avenue. Before Greenport was detached, there was an area around the compact part of Hudson, which permitted taxes to be levied for improvements in the city proper. Farmers would not derive much benefit. I think that the creation of Greenport was to replace the compact area, with very few farms in area.

It was when the east-west boundary was created in 1855, that the north-south boundary was changed to be 3rd Street extended. The southern part of S 3rd Street hasn't always been known as 3rd Street, and it may swing just a tad bit more east to get around South Bay.

[MASHED POTATOES. VOTE!]

I really like the council's meeting room. Sort of cross between a small legislative chamber (like VT Senate) and an university classroom.

[jimrtex]

Fair Apportionment and the Banzhaf Index(PDF)

This is an interesting article, given its author, Bernard Grofman, who some will recognize as the special master in the Virginia congressional redistricting case.

The Banzhaf Power Index as used in New York is based on the probability that a given representative in a legislative body will cast a decisive vote in a political body based on the following assumptions:

(1) All representatives are equally likely to vote Aye or No.
(2) All combinations (coalitions) of representatives are possible and equally likely.

A vote cast by a representative is decisive if

(1) The coalition has a majority weighted vote (or meets some other threshold such as a 2/3 supermajority);
(2) Switching the vote of the representative alone to No, causes the coalition to fail (to achieve a majority).

When a coalition fails to achieve a majority, no votes are decisive, since they have already lost, and any vote-switchers to No will only make it worse.
When a coalition achieves a substantial majority, no votes are decisive, since even with one defector, the coalition will have a majority.

When the vote is fairly close, the votes of some representatives will be decisive, while that of others will not. For example, imagine that a coalition achieves 53% of the vote. A representative with 4% of the vote will be decisive, since if he abandons the coalition it will have 49% of the vote. If a representative with 1% of the vote leaves the coalition it will make no difference.

The Banzhaf Power Index may overestimate the power of representatives with greater weight. Consider the previous example, and imagine two representatives with 2% of the vote each. Individually, their vote was not decisive. But collectively, the vote of at least one of them was _necessary_ to achieve a majority, assuming that no other representatives changed their vote.

Banzhaf later extended his index to voting in popular elections. Since all voters have one vote, a deciding vote can only occur when there is a one-vote victory. While exceedingly unlikely when there are a substantial number of voters, it is not impossible. While computation of the BPI requires enumeration in the general case; it can be determined analytically when all voters have the same weight.

It turns out that the likelihood of a one-vote victory is inversely proportional to the square root of the number of voters, rather than inversely proportional.

His theory was considered in Whitcomb v Chavis. This Supreme Court decision in 1971 considered whether elections to the Indiana legislature from multi-member districts violated One Man, One Vote. At the time Marion County (Indianapolis) had 15 representatives and 8 senators elected at large. The thrust of the complaint was that the black area of Indianapolis was denied representation. A second part of the complaint was based on the idea that permitting Marion County to elect 15 representatives at-large was discriminatory to voters in other districts with fewer representatives. In particular, Marion County was compared to Lake County (Gary, Hammond). The two counties had 15 and 11 representatives respectively, which happened to be quite proportional to population. The ratio of population was 1.36:1.00, but based on Banzhaf's theory the representation ratio should have been 1.17:1.00. Actually, the argument made was that it was impossible to reconcile Banzhaf's theory with anything other than equal population districts.

The district court (3-judge panel) ruled that not providing single-member representation to Marion County was unconstitutional and then proceeded to order statewide redistricting, but not based on Banzhaf logic, but on the basis of malapportionment. Indiana had successfully redistricted post-Reynolds in 1965, and had the result approved by a federal district court. But by 1969, the Supreme Court had indicated that the deviations of 28.2% for the senate, and 24.8% for the house were no longer tenable (in the district court decision, the court calculated the number of districts that had greater than 1% deviation, suggesting that they were demanding almost precise equality).

The Indiana legislature had already met in 1969, and did not have a 1970 session, so the court imposed its plan to be used for the 1970 election. The Supreme Court stayed their decision. Oral arguments were held in December 1970, after an election conducted using the 1965 post-Reynolds boundaries.

In the January 1971 session, the newly elected legislature passed a single-member district plan that violated the Indiana Constitution provision on the division of counties. This was noted by the Supreme Court when it finally issued its opinion in June.

The Supreme Court ruled that Marion County did not have to be divided into single member districts, but that because the district court had ruled what it did, it opened up the state to redistricting
It appears that the fact that there would have been a partial redistricting opened up the state to a total redistricting based on deviations outside of Marion County. But because of the Supreme Court stay those boundaries were never used, and the last conceivable election had been held before they heard the case. By the time they had decided the case, the legislature had redistricted on the basis of the 1970 Census.

There was a 5-member majority that ruled that Marion County did not have to be divided. Justice Potter Steward did not support the conclusion that statewide redistricting was nonetheless required. A 3-member dissent said that Marion County should be split, and that statewide redistricting was proper.
Justice Harlan filed a dissent.

Justice Harlan lampooned the Banzhaf theory, and pointed out how the probability of "decisive" votes would be degraded with minor changes to the base assumptions. He considered an election with 300,000 voters (Marion County had about 700,000 persons at the time). If voters used loaded coins that came up heads 50.5% of the time, there would be an astronomical decline in the probability of a one-vote election. Alternatively if 10,000 persons voted one way, and 5,000 the other, adamantly ignoring their coin flips, while the other 285,000 citizen-voters dutifully flipped their coins, there would be an even greater decline.

It was these claims by Harlan that Grofman investigated. Since Grofman wrote about a decade later, presumably Harlan had read the numbers in a brief and added them to his opinion.

The majority didn't appear to really understand the concept of the Banzhaf theory as applied to large numbers of voters. Their example posited an example of a county with nine (sic) voters and three representatives, vs. three districts with three (sic) voters each. They spent a lot of time discussing the situation in Marion County, and then jumped into the disparity in population among districts statewide, with no acceptance of the Banzhaf theory. They concluded that even though the district court erred in dividing Marion County, that once they did they could redistrict the entire state.

The other 3-member dissent said that Marion County should have been divided, and the state redistricted, did not mention Banzhaf at all.