Note widespread use of Proportional Representation in the US: presidential primaries, convention delegate selection!


Many of us have been pushing for more widespread adoption of election methods that promote Proportional Representation (PR), to make more votes count, to produce more representative elected bodies, to eliminate or reduce gerrymandering, and for many other reasons.

Most voters in the US are unfamiliar with PR, but they shouldn’t be. I’m surprised that we don’t hear more about it since so many of us use it every 4 years. I’m talking about nearly all the Democratic presidential primaries, and many of the Republican primaries, about which The Green Papers Election Glossary has this to say:

Seeing the WINNER-TAKE-ALL primary as unfairly reducing the input of significant minority factions within the party in the presidential nominating process, the McGovern-Fraser reforms of the early-to-mid 1970’s successfully promoted the so-called “PROPORTIONAL” type of primary as an alternative to be used in the Democratic Party’s nomination process. In the PROPORTIONAL type of presidential preference primary, the district delegates are apportioned among the top vote-getters in each (usually congressional, but occasionally state legislative) district while the at-large delegates are apportioned among the top vote-getters statewide by the percentage of the vote received above a certain threshold (most often 15 percent: a figure actually mandated by the rules of the Democratic Party since 1992). This is the system used by the vast majority of the states holding presidential primaries in the Democratic Party; the Republican party (where WINNER-TAKE-ALL primaries are still permitted) uses it in far fewer states than the Democrats and, in the vast majority of these, the GOP usually started using the PROPORTIONAL type only because Democrat-dominated State Legislatures of the mid-to-late 1970’s passed laws forcing both parties to use this type of presidential preference primary. The major difference between the two parties’ PROPORTIONAL primaries is in the thresholds used by the Republicans, which can vary from as much as 20 percent or more to as little as virtually 0 percent. (as noted below, the Democrats are currently required by party rules to use a 15 percent threshold in all their PROPORTIONAL primaries).

Such proportional allocation methods are also used starting at the grassroots in many other party settings, e.g. for seating delegates in county and state conventions based on caucuses.

I’d love to see this experience and history of PR in the US fleshed out, and figure out good ways to use it to inform our discussions of good PR methods, and for advocacy for PR. Has anyone seen that sort of documentation or discussion?

  • Can anyone point to a more detailed history of PR methods by parties?
  • Does anyone have good easy-to-parse data feeds for how these allocations have been made? It looks like the GreenPapers have lots of detailed data, like Colorado Democratic Delegation 2016, but I don’t know if they have it all organized for easy download and analysis.
  • Does anyone know of a concise encoding of the exact methods used in each jurisdiction for each election? Again, Green Papers has lots of detail (The Math Behind the Democratic Delegate Allocation - 2016), but I’d love to see code that could reproduce all the historical allocation calculations based on the data.

I’ll briefly note that this also has a variety of implications for risk-limiting audits of elections, which Colorado has been pioneering recently. If the goal is to audit whether the detailed outcome of the election is correct, i.e. the full delegate allocation, the effective margins become very close even when it is very clear which candidate won the state. For related calculations which can be easily reproduced online for free, see IPython notebook and paper with Risk-limiting audit code for Proportional Representation via Highest Averages. That’s one reason for wanting detailed, easily-parsed data.

Many thanks,

Neal McBurnett -
Center for Election Science Board member


At least in the U.S., there is fairly strong support for the concepts of freedom of association and freedom of speech. This brings the consequence that any group of individuals may start a new party, which may select and support one or more political candidates. And no one can create laws or regulations prescribing the manner in which they must carry that out.

The only reason why people might even consider the potential for such laws and rules is the existence of two-party (or even just few-party) lock-in, which is enabled by defective electoral systems.

If we were to utilize a simple score voting system (hopefully with hand counted paper ballots, and widespread usage of the strategic hedge strategy by the voters), we could completely disrupt the two-party lock-in system.

Since political parties are essentially private organizations, proportional representation may only be required to apply to official general elections.

If we use terms such as “winner take all” we then seem to be begging the question of “who should take the rest” (otherwise)? And people may answer this question in many different ways.