Helsinki citizens assembly: the results are in!

A more democratic type of referendum

By Metta Spencer

The Structure Committee of HCA is supposed to develop fair, democratic decision-making processes for that organization. To that end I have been experimenting with a more democratic type of referendum. (See the article above.) In the past several issues of Peace there have been opportunities for our readers to participate in this experiment. To enlarge the numbers, I asked my sociology students to participate. Now the results are in and I will explain what it's about.

Voters do not care equally about all issues, and the votes of those who do not care about a particular issue cancel out the votes of those who do care. It would be better to cast several votes on the issues we care about and no votes on the issues we don't care about. That would be fair, if all voters may "spend" the same number of votes however they please. Though this arrangement benefits everyone, it should especially please groups who are always outvoted, even on issues that are dearest to them. With this weighting scheme, the intensity of their wishes can partly make up for their small numbers as voters.

Sometimes ethnic groups that are minorities in the large nation-state constitute local majorities. By taking their republic out of the large state, they can change from being a minority (always outvoted) to being a majority (always winning elections). However, there are grave drawbacks to this remedy. In view of the frequency with which secession leads to warfare and economic hardship, peace activists should do everything possible to make secession unnecessary and unattractive. This requires us to create opportunities for minorities to win-without estranging the majority. That is the objective of our experiment with the referendum. We want to allow ethnic minorities to form powerful voting blocs without leaving the larger state and without fighting over territory. Ethnic loyalty, when coupled with a claim for a separate political territory, is called nationalism. If ethnic identity can be expressed politically and effectively in some other way, it will be unnecessary to make nationalistic claims.

Historically, ethnicity tends to wax and wane. It is undesirable to lock it in by making it the basis for fixed, sovereign nation-state boundaries. Let us make ethnicity a voluntary matter, allowing people to form constituencies by registering, instead of grouping them by district of residence. Then whenever one changes her group identity, she can simply change her registration to the new preference.

Here is how our referendum worked:

Stage One. I asked each participant to name three of their group identities. The most frequently chosen of these identities would become voting constituencies-a "riding" that is defined, not geographically, but in terms of social identification. Also, I asked everyone to suggest a proposition for a Canadian referendum; if it passed, hypothetically it would become law.

In tabulating these responses, I found that nine different social identities were chosen frequently enough to warrant becoming voting constituencies. Two of these were ethnic identities (Anglophones and Asians), one was religious (Catholics), two were defined by gender (males and females), and the rest were social functions (environmentalists, feminists, peace activists, and students). But this is Canada. In some other societies where ethnicity is salient, such as Sri Lanka or Georgia, ethnic groups might be the only constituencies.

Next I examined the array of possible propositions that had been suggested by each constituency. I identified the two topics that had been named most often. It was necessary to re-word all potential propositions somewhat, since the original authors had phrased them in varying ways. Stage Two: I asked participants to choose one of the nine constituencies and also to choose which of the two propositions posed by their constituency should be placed on the referendum.

Stage Three: I asked participants to vote on the referendum itself. There were nine propositions, and each voter would therefore cast nine votes. Unlike ordinary referendums, however, the nine votes could be distributed across the entire array of items to indicate intensity, as well as direction, of opinion. The ballot printed in Peace asked readers to vote in an unusual way. Instead of voting yes or no on each of the propositions, each voter was asked to cast nine votes, but distribute them as he or she preferred across all nine propositions. One voter might cast all nine votes against proposition 3. Another voter might cast three votes in favor of propositions 1,5, and 7. A third voter might cast one vote on propositions 1,2,3,4,5, and 6, and two votes on proposition 8, not voting on proposition 9. The possibilities are endless.

In my classes, I asked students to vote in the same way, but also to vote in the conventional way-one vote per proposition. I then compared the results of the two systems, which are shown below.

1. Canada shall increase support for sustainable development at home and around the world.

Conventional system: 144 yes, 15 no (91% yes)

Experimental system: 115 yes, 4 no (97% yes)

2. Tobacco may not be sold in Canada or smoked in public places.Conventional system: 77 yes, 81 no (49% yes)

Experimental system: 79 yes, 42 no (65% yes)

3. The Young Offenders Act shall be repealed and stronger penalties shall be imposed on juveniles who commit crimes.

Conventional system: 143 yes, 11 no (93% yes)

Experimental system: 204 yes, 15 no (94% yes)

4. Manufacturers must immediately cease using chloroflurocarbons in Canada. Conventional system: 149 yes, 11 no (93% yes)

Experimental system: 149 yes, 2 no (99% yes)

5. Women who wish to care for their own children at home may receive a subsidy so they can do so instead of working for pay elsewhere. Conventional system: 104 yes, 55 no (65% yes)

Experimental system: 109 yes, 21 no (84% yes)

6. The principle of "equal pay for work of equal value" shall be enforced for all jobs in firms with 10 or more employees.

Conventional system: 143 yes, 15 no (90% yes)

Experimental system: 208 yes, 8 no (96% yes)

7. The use of marijuana and other psychotropic drugs shall be legal in Canada. Those who are addicted to narcotics shall be provided doses under close medical supervision.

Conventional system: 54 yes, 102 no (35% yes)

Experimental system: 33 yes, 58 no (38% yes)

8. Arms shall not be exported to any country that is known to have threatened human rights within the preceding three years.

Conventional system: 151 yes, 5 no (97% yes)

Experimental system: 162 yes, 8 no (95% yes)

9. The quality of Canadian universities shall be improved and grants and loans provided to low-income students in all post-secondary institutions.

Conventional system: 148 yes, 12 no (92% yes)

Experimental system: 231 yes, 5 no (98% yes)


In few cases are there large differences between the results of the conventional and experimental systems. The two most interesting comparisons are items 2 and 5. In the case of Proposition Two, the outcome is actually reversed. In a conventional referendum, the measure would narrowly be defeated. In the experimental referendum it carries by a sizeable margin, 65% to 35%. This result is possible because here a number of voters do not have strong opinions and opt to allow those who do have strong opinions make the decision. If this were a legally binding referendum, it would become illegal to sell tobacco in Canada or to smoke it in public places! Propsition Five, which would subsidize child care at home, carried in both the conventional and experimental systems. (Of course, if the referendum had been real, the public debate might have scared the electorate into voting against it because of its high cost.) The interesting relevation gained by comparing the two outcomes is that many people chose to sit on the sidelines and allow the decision to be made by those who cared strongly about the matter. Those who did vote in the experimental system were much more favorable to the measure than in the conventional system. Of course, we are not interested in these substantive issues, but only in the method that they illustrate. However, if the referendum results were based on a full electorate, it would mean that in the issue of subsidized home child care, politicians could afford to be bold. The opposition would not be very intensive, while those who favor the measure would be extremely supportive. What does this experiment mean for those states facing nationalistic secession movements? It provides a mechanism for forming ethnic constituencies that can endure as long as ethnicity is salient-without carving up territory or fighting over boundaries. Ethnic communities that are perpetually outvoted under conventional democratic processes would have a chance to win on certain issues, without needing to secede. And when the minority did win, it would be with the acquiescence of the majority, for had they not acquiesced, they would have voted against the measure and blocked it. The key to keeping multi-cultural states integrated is to find and make the most of such areas of consensus.

If you know of an organization that may benefit from this model, please propose it. We need to prompt real organizations to adopt this approach. Only after its success has been demonstrated in NGOs will nations consider adopting it.

Peace Magazine Jun-Jul 1993

Peace Magazine Jun-Jul 1993, page 25. Some rights reserved.

Search for other articles by Metta Spencer here

Peace Magazine homepage