Partitioning the State: Some Basic Observations
by
Pierre Lemieux
by
Pierre Lemieux
Debates on partitioning the Québec
territory after an eventual secession neglect a few basic facts.
Consider the following simple model. A continuous territory called Q is attached to a State called C. Q is divided in three regions: M (for, say, "Montréal"), S (for "Saguenay") and E ("Eastern Townships"). Suppose now that 51% ("the Whites") of the 400 residents of Q want to secede from C, and 49% ("the Blacks") do not. Numbers vary according to regions, as given in Table 1 (neglect for now the number of residents in county subdivisions). Figure 2 provides a stylized map of our simplified model. Regions where the majority wants to stay with C are black; the region where the majority wants to secede is white.
Given our assumptions, the majority (51%) of Q votes for secession. Assuming partition by region is feasible, the majority of both M and E will vote to secede from Q. The territory will be partitioned according to the pattern shown in Figure 2. Call this "Partition A".
The problem is that, even in our simple model,
this is not the only way to partition Q when one considers
subdivisions (say, counties) within each arbitrary region. Figure
3 pictures another possible partition, on the basis of county
voting figures of Table 1. In this partition ("Partition
B"), bold numbers show the White votes, and roman numbers
the Black votes, for each county. What happened is that County
{18, 16} and County {20, 19} of Region E
have voted to go with Q, while County {20, 25}
of region S has voted to remain with C.
Nor is this the end of the matter, for other
partitions are possible. In "Partition C" (Figure 4),
County {35, 25} of Region M has voted to be part
of Q in order not to be isolated in C. This has
the effect of isolating County {20, 25} of S --
given the constraint of continuous territories.
The first fact illustrated by this simple numerical example is that there is no unique solution for continuously partitioning a territory: the solution depends on the units chosen for the partitions. If we moved from county-level units to still smaller units (say, the municipal level), the number of possible partitions would increase.
The total proportion of satisfied individuals
(i.e., of those being in the territory they have voted for) can
be increased by refining the partitioning. As this happens, however,
the proportion of dissatisfied minority voters will also increase.
This can be seen in Figure 5. Starting from the status quo, block
secession of Q increases the proportion of satisfied Whites
(to 100% in this limit case), but diminishes the proportion of
satisfied Blacks (to 0% at the limit). When post-secession partitioning
starts, the Whites' satisfaction decreases as the Black's climbs.
The Whites will fight partitioning for the very same reason that
the Blacks opposed the original secession. But when partitioning
has started, any new move will work in favor of one group and
against the other.
A second fact then emerges from our analysis. Since different partitions will change the proportion of individuals satisfied in either group, the solution chosen among the many possible ones will depend on each group's political power (which is ultimately based on force). The political partitioning process is, as economists say, only a matter of shifting gains and losses among groups.
Allowing individuals to move will minimize dissatisfaction, but moving costs will be borne by the less politically powerful, i.e., the ones who have not succeeded in redrawing the partitions in their favor.
Thirdly, as long as the regions were not homogeneous to begin with, there is no level of continuous territorial partitioning, no way of drawing political borders, that can make every group 100% satisfied. The only way to do this would be to dispense altogether with the idea of continuous territories, i.e., of state territorial monopolies, and let each individual (the smallest group possible) decide which state he will belong to -- or, for that matter, if he will belong to any at all.
A final observation can be made. Short of anarchy or discontinuous territorial States, the advantage of a minimal state is to make it easier for Whites and Blacks to live together. The more policies a White or a Black majority wants to impose in its State, the more oppressed the minority will be. Secessions and partitions come from the desire of majorities to impose policies that minorities do not want.