FROM TIM EDWARDS (rec'd Oct. 27, 2007)

Is Thomas Gangale’s “American Plan” for revamping the U.S. primary system a good plan, or are there serious flaws we should address before we go through the time and effort to change the system?  Let’s investigate.


Gangale’s original plan, before his later modifications, was built on two key concepts:

INCREASING SIZE:  Primaries start small and steadily increase in size to allow less-funded candidates to compete.

RANDOM SELECTION:  Random selection of states into groups of primaries to ensure fairness and to prevent camping out years in advance.

These goals are good ones, and they should apply to any primary system we put into place.  But as we will see below, the American Plan has serious underlying problems.



(1)  Too many primaries.

A series of 10 primary groups is too long for a primary season.  The other main proposals being discussed (the NASS’s Rotating Regional Primary Plan and the Delaware Plan) both have only early historical primaries, plus 4 additional groups.

(2)  Abandonment of “increasing size”, and irrelevant late, small primaries.

In order to accommodate California, Gangale changed the schedule (originally 8,16,24,32,40,48,56,64,72,80) to a staggered one (8,16,24,56,32,64,40,72,48,80).  This creates small, late primaries.  And more importantly, under the American Plan, the last 3 primaries occur after half the delegates have been awarded.

(3)   High travel costs.

The original goal of the small initial primaries was to allow less-funded candidates to compete with those who have more funds.  This is not possible if the first block of primaries includes Hawaii , North Dakota , Mississippi , and Vermont , as it well could.

(4)  Geographic outliers ignored/No use of “strength in numbers” for small states..

If the first primary includes Nevada , Utah , Wyoming and Vermont , few candidates will spend time in Vermont ..  Small states gain much more relevance as part of a “pod” of nearby, similar states.  This allows candidates to address regional issues, as well.

(5)  “Cliff” advantages/disadvantages to certain-sized states, and resulting liberal political bias.

Gangale’s plan treats 8-district states much differently than 9-district states, even though they are very similar in size.  This early advantage is even more pronounced for Michigan , which with 15 districts is the largest state that can fit into the 16-district primary, and is the single state most likely to be in the first 3 primaries.  Although the first 3 primaries only constitute 11% of congressional districts, Michigan alone has a 24% chance of being in the first 3.

When you chart this cliff effect by the political leaning of the states, you get an astounding result:  the American Plan has a significant liberal bias.  The states with the most benefit in the first four groups, the 8-district primary (MD, WI, MN, AZ), the 16-district primary (MI), the 24-district primary (IL, PA), and the 56-district primary (CA) include 7 blue states, but only 1 red state.


*** CAN IT BE FIXED?? ***

The American Plan would be much easier to fix if all the states had the same number of congressional districts.  There would be no cliff effects, and every state would have the same fair shot at going early or late.

Of course, this isn’t possible.  Or is it?  Can we group all the states together with politically-similar, contiguous states, into same-sized buckets?

Not quite, but we can get pretty close.  Let’s take the largest states off the table first.  We’ll put California , Texas , New York , and Florida aside (and with them the group of historically early states, Iowa , New Hampshire , Nevada , and South Carolina ) for now.

The next seven largest states all have 16 districts, plus or minus 3.  (These include IL, PA, OH, MI, NJ, NC, and GA.)  Can we combine smaller states to have about 16 districts, as well?  Our goal should be to combine politically-similar, contiguous states.  Indiana and Kentucky are both more conservative than their neighbors, and they combine to form a 15-district pod.  Minnesota and Wisconsin have a similar liberal/independent voting population, and together they form an 18-district pod.

When you continue this through the rest of the country, you end up with 19 pods of approximately 15 or 16 districts.  Now let’s go back to the large states.  California has about 3 times that number of districts, and the other three (TX, NY, FL) all have about twice that number.  Let’s just give each pod 1 point for every roughly 16 districts to simplify, and we have:

California :  3 points
Texas , New York , Florida : 2 points each
The other 19 pods:  1 point each

Total:  28 points for the country.

Randomly assigning primaries, with a few rules to keep it politically unbiased, we can use the following schedule:

Preliminary states (IA, NH, NV, SC)
First primary:  1 point  (about 3% of the population)
Second Primary:  2 points  (about 7% of the population)
Third Primary:  4 points  (about 14% of the population)
Fourth Primary:  7 points  (about 24% of the population)
Fifth Primary:  14 points  (about 48% of the population)

Details on the rules to make it politically unbiased are at

under the “Modified American Plan” section.



Let’s go back to our “Five Fatal Flaws” and check:

(1)  Too many primaries.

We have cut from 10 primary dates to 5 (plus early states), a significant reduction.

(2)  Abandonment of “steadily increasing”, and irrelevant late, small primaries.

The primaries are now steadily increasing, and there are no irrelevant groups.

(3)   High travel costs.

Under the Modified American Plan, the first primary is always contiguous.  The second primary is contiguous just over 50% of the time, and even if not, it is in only 2 contiguous, roughly equal-sized groups.

(4)  Geographic outliers ignored/No use of “strength in numbers” for small states.

All small states are merged in with their neighbors.  Delaware always votes with Maryland and DC, Vermont and Rhode Island always vote with the rest of New England , and so on.

(5)  “Cliff” advantages/disadvantages to certain-sized states, and resulting liberal political bias.

Cliff advantages and disadvantages have been removed, since we are no longer sticklers for the exact size of a state or a pod.  The four largest states have slightly different rules, but since the assignment rules are different for the Modified American Plan (see, the political skew is removed.



If we’re going to go through the effort of changing our primary calendar, let’s not change it to another flawed system.  Let’s do it right, and consider ALL the important factors, including travel time, political balance, and strength in numbers for small states.

(For more information on the American Plan, the Delaware Plan, the Rotating Regional Primary Plan, and several modifications of the above, please check out the detailed analyses available at:

The various plans have been graded on a fair scorecard, and compared against each other.)

Let’s not go for the solution that’s the most-hyped.  Let’s go for one that’s RIGHT for America .

Thank you,

Tim Edwards