James Green-Armytage
Collective Budget-Setting Proposal
In some
cases it may be interesting to engage in a collective budget setting process,
where everyone controls an equal share of a government budget, and is able to
allocate that money to the project of their choice.
Choices of
different projects should probably still be bounded by the legislature in some
way. That is, you wouldn't be able to spend the money on absolutely anything
you want, such as buying yourself a car or whatever... Rather, there should be
a list of approved projects that you could choose from. Perhaps it would make
sense to require something like a majority of the legislature to approve a
project to begin with.
You
wouldn't need to determine an entire budget this way. If you like, you could
determine a certain portion of the budget this way and then have the
legislature fill in the rest on its own.
It makes
sense to use STV as the basic method to translate the individual allocations
into the collective allocation. Projects that are overfunded could have the
surplus money transferred to people's second choices; projects that are too
severely underfunded to be effective might also transfer funds to people's
subsequent choices. If possible it makes sense to use one of the more
sophisticated forms of STV here, such as Meek or CPO-STV.
You could
also allow people to divide up 'their' allocation money so that it all doesn't
go to the same place. For example, if I have $10,000 in allocation money, I
could say that $6,000 goes to project A, and then if project A is overfunded,
to project B, and then to project C, and so on... and meanwhile also say that
the remaining $4,000 goes to project R, to project S, to project T, and so on.
However,
this leaves somewhat open the issue of how much money each project should receive
before additional money is transferred elsewhere. What if someone would like to
contribute to project A, but they believe that the project cap is set too high,
and while they'd like to fund A up to a point, they wouldn't like to fund it
all the way to its official cap.
So what
I'm trying to do here is to devise an effective system where people can decide
the level of funding which a given project should reach before the remainder of
their money is transferred elsewhere. At first I came up with some inferior
methods of doing this, but I won’t bore you with those right now. Here is what
I propose:
At each
amount of a project's budget, ask how many people are willing to contribute,
taking into account that the amount will be divided by the number of contributing
people. There are a couple factors that determine whether someone is willing to
contribute up to that amount.
One factor
is whether the total amount that I have allocated for the project is sufficient
to pay my share of the total amount. For example, let's say that there are 20
contributors for project X, including myself. Let's say that I allocate $500,
and everyone else allocates $600. The total amount for the project would
gradually climb from zero upwards, asking with each increase whether everyone
is still willing to contribute at that level. When the total amount is $2,000,
then the share for each of us is $2,000 / 20 = $200, and yes, we are all
willing to contribute $200, so the amount keeps increasing without
interruption.
The amount
climbs without any change until it reaches $10,000, at which point our
individual share of the cost is each $10,000 / 20 = $500. Now we have reached
the highest point at which I am willing to pay. At $10,001, my share would be
greater than $500, which I did not agree to. So I contribute my $500, everyone
else contributes their $500, making a total of $10,000. They each still have
$100 to spend on this project, but I'm out, and so now when the amount starts
climbing again, I'm not a part of it, although the rest of them still are. At
$10,001 they're all still in, at $10,002 they're all still in, and so on, up to
$11,900, which is the point at which they've all spent their remaining $100 and
have nothing left to give. So the budget for the project (so far, at least) is
$11,900.
The other
factor is whether the amount is greater than the cap I have indicated for that
project. For example, let’s say that I am willing to allocate $500 to project
Y, but only up to a project cap of $5,000, whereas there are 19 other people
who are willing to allocate $500 with a project cap of $6,000.
At a total
budget of $1 we would all be in, same with $2, $3, and so on up to $5,000. When
we get to a budget of $5,000, we’ve each contributed $250. So, I contribute my
$250, but from that point on I’m no longer willing to contribute anything. So
the remaining $250 that I would have been willing to allocate to the project
under different circumstances (i.e. less contribution from others) is instead
moved along to whatever project I have indicated as the second priority for
that money.
However,
the budget for project Y keeps climbing due to other people’s contribution.
Everyone else is still in at $5,001, at $5,002, and so on up to $6,000. The
extra $1,000 is divided 19 ways, so everyone else pays another $52.63
(approximately). Added to their prior contribution of $250, that’s $302.63 for
each of them, and $250 for me, with the project budget now at $6,000.
So, what
happens to the extra $250 of mine that was transferred along to another project?
Does it just get added to the budget of that project after all the other
calculations have been made, and stick there? Well, no. What you would actually
want to do is start a whole new round of calculations, for all the projects at
once, taking into account the new money that was transferred into them from
other projects. That money will enter into the calculation for its second
choice project on an equal basis with other any other money offered for that
project, whether first choice or second choice or whatever. I may have
specified a project cap to go with it into the second project, or I may not
have. The rules are the same as before.
At the end
of the second round of calculation there may once again be surplus money from
various projects. If so, then a third round will begin, and so on. Eventually
this process will come to rest on its own accord.
Note: In
addition to people’s individual caps for a particular project, the some
projects could also have a total cap of their own. Anyone giving to that
project who doesn’t indicate a cap, or who indicates a cap higher than the
official cap, is treated as if their spending cap is equal to the official cap.
Here is a
more detailed example. There are 9 people (A, B, C, etc.) willing to allocate
various amounts of money to a given project. (For example, A allocates 350, B
allocates 250, and so on.) The project itself has no set cap, but the different
contributors have indicated various caps. (For example, A indicates a cap of
1200, G indicates a cap of 1000, etc.) You can imagine that the figures are in
dollars, or whatever currency you like. Here is the calculation:
A B C D E F G H I Project
cap
1200 1200 1200
1200 1200
1200 1000
950 1000
none
amount 350 250
150 60
50 40
150 200
500
9 up to 360
-40 -40
-40 -40
-40 -40
-40 -40
-40 +360
310
210 110
20
10
0 110
160 460
360
8 up to 440
-10 -10
-10 -10
-10
-10 -10
-10 +80
300
200 100
10 0
100
150 450 440
7 up to 510
-10 -10
-10 -10
-10 -10
-10 +70
290
190
90 0
90
140 440
510
6 up to 950 -73.3 -73.3 -73.3
-73.3
-73.3 -73.3 +440
216.7 116.7
16.7
16.7
66.7 366.7 950
5 up to 1000
-10 -10
-10
-10
-10
+50
206.7 106.7
6.7
6.7
356.7 1000
3 up to 1020 -6.7 -6.7
-6.7
+20
200
100 0
1020
2 up to 1200 -90 -90
+180
110
10
1200
contribution
240 240
150 60
50 40
43.3 133.3 143.3
remainder
110 10 0 0 0 0
6.7 66.7
356.7
Notice
that at each stage of the calculation everyone who is still involved is paying
the same amount of money. Some people stop contributing because they have
exhausted the amount of money that they were willing to contribute (in this
example F, E, D, and C). Some people stop contributing at the point where they
feel the project has sufficient funding (in this example H, G, I, and
eventually A and B as well).
The end
result is that the project has a budget of 1200, and each person contributes
the amount shown above. The remainders shown above are passed along to whatever
project those people indicate as the next priority for that particular gob of money.
Is collective budget setting a good idea? I don't know. But if it is done, I recommend this algorithm as being the best way to do it, assuming that the needed computational resources are available.