decodedscience
  • オンラインカジノ
  • 私たちについて

How to Divide Indivisible Goods Fairly: Algorithm for Dividing Assets

February 23, 2014 by Mike DeHaan Leave a Comment

Share9
+11
Share4
Pin
Tweet
Shares 14

How do you divide assets when there’s conflict over ownership? Math always finds a way. Image by xenia

We know how to allocate items fairly if the goods can be divided into equal portions: share and share alike. Divide the pie into equal-sized slices, with one slice per guest. Liquidate the assets of an estate and share the proceeds equally with each inheritor. Live in a time-share vacation property for the same number of weeks per year.

Indivisible goods, however, are indivisible. Examples include: real estate when joint ownership is excluded, for one example. Is there a way to ensure an “optimal envy-free allocation” between two parties?

Steven J. Brams, D. Marc Kilgour, and Christian Klamler recently wrote Two-Person Fair Division of Indivisible Items: An Efficient, Envy-Free Algorithm that provides two solutions.

Here we explain their stronger method, the “AL algorithm,” with some comments from one of the authors.




How to Divide an Inheritance Fairly

Suppose that Alice and Bob’s parents die. The will allows the two adult children to divide the inheritance between themselves in any way they see fit. There are a number of items, such as houses and cottages, cars and boats, and one piece of antique furniture. Both Alice and Bob would prefer to keep most items “in the family” rather than liquidate these assets; but they refuse to consider joint ownership. They agree to work with a mediator, and that a fair allocation would satisfy certain conditions:

  • Alice and Bob will each receive the same number of items.
  • Some items might not be assigned to either Alice or Bob; these “contested” items need further consideration outside of this algorithm.
  • Alice and Bob will each submit their own prioritized list of the items to the mediator.
  • After the allocation, the mediator will make a list for Alice, and a list for Bob, of the pairs of items allocated to each of them. Alice’s list will show she preferred her n-th item over Bob’s n-th item; likewise Bob’s list will show his pair-wise preference for his items. (This is the “envy-free” promise: to prefer one’s n-th item over the other party’s n-th item).

The paper by Brams, Kilgour and Klamler gives a mathematical proof that their algorithm will create this kind of fair allocation. The proof is too long for this article; instead, let’s focus on how the algorithm works.

How to Prepare for the Allocation Algorithm

An algorithm is a set of rules, which may be followed by a computer or by a person such as the mediator in this example.

Alice and Bob must each prepare their list of items in order of their preferences. For example, Alice might list the New York townhouse, then the vacation property in Bermuda, the yacht, the detached home in San Francisco, the Arizona ranch, the BMW, the antique dresser, and finally the pick-up truck. Bob may prefer the detached home, the ranch, the townhouse, the vacation property, the pick-up truck, the BMW, the yacht and finally the dresser. These preferences may reflect their different lifestyles, sentimental attachments, or simply the cash value of each item.

For simplicity, let’s say that Alice has an ordered list {1, 2, 3, 4, 5, 6, 7, 8}. The mediator notes that #1 corresponds to the New York townhouse, and so on. Bob’s list for this first example is {3, 4, 7, 1, 2, 6, 8, 5}.

Allocation Preferences Example 1: How will the algorithm allocate these resources? Image by Mike DeHaan

The paper calls this the AL algorithm. It begins by setting the Stage# and the # of Pairs Allotted to zero. It also begins with an empty list of items allocated to Alice, to Bob, as well as an empty list of Contested items. Then the algorithm receives the two lists of preferences. Then the algorithm follows these rules:The AL Allocation Algorithm

  1. What is Alice’s preferred unallocated item? If nothing remains, stop.
  2. Is Bob’s preferred unallocated item different from Alice’s? If they are the same, go to step 3. If they are different, then:
    • Allocate Alice’s item to Alice, and remove it from Bob’s list.
    • Allocate Bob’s item to Bob, and remove it from Alice’s list.
    • Add one to the # of Pairs Allotted.
    • Return to step 1.
  3. Alice and Bob want the same “tied item”. If the # of Pairs Allotted equals zero, then go to step 5. If there are no further items on Alice’s list (or Bob’s list), then go to step 5. Otherwise:
    • List Alice’s unallocated items that she ranked below the tied item, in order of her preference. This is list AU (“Alice’s Unallocated” list).
    • List Bob’s unallocated items that he ranked below the tied item, in order of his preference. This is list BU.
    • Consider allocating the tied item to Alice, and then go through the BU list for Bob. For each item in the BU list:
      • If Bob gets this compensation item, count the number of items that Alice has received plus those that remain unallocated, that Bob would have preferred to this compensation item. If this number is greater than the # of Pairs Allotted, then the current compensation item is not feasible; otherwise we can allocate this compensation item in step 4.
    • Consider allocating the tied item to Bob, and then go through the AU list for Alice. For each item in the AU list:
      • If Alice gets this compensation item, count the number of items that Bob has received plus those that remain unallocated, that Alice would have preferred to this compensation item. If this number is greater than the # of Pairs Allotted, then the current compensation item is not feasible; otherwise we can allocate this compensation item in step 4.
    • If no compensation item allocation was feasible, go to step 5.
  4. Much like the successful step 2: Remove the tied item from Alice’s list and from Bob’s list, and place it onto the Contested List. Return to step 1.
    • Allocate Alice’s item to Alice, and remove it from Bob’s list.
    • Allocate Bob’s item to Bob, and remove it from Alice’s list.
    • Add one to the # of Pairs Allotted.
    • Return to step 1.
  5. Remove the tied item from Alice’s list and from Bob’s list, and place it onto the Contested List. Return to step 1.

The article presents a mathematical proof that this algorithm will provide one of the best envy-free allocations for any list that Alice and Bob can present. There may be several equally good allocations; or everything might go to the Contested List. However, this algorithm will provide one of the best possible allocations.

Click to Read Page Two: Allotting the Estate

© Copyright 2014 Mike DeHaan, All rights Reserved. Written For: Decoded Science
Share9
+11
Share4
Pin
Tweet
Shares 14
Pages: 1 2

Filed Under: Calculations Tagged With: indivisible goods, mathematics

Resources for this article



Decoded Everything is a non-profit corporation, dependent on donations from readers like you. Donate now! Your support keeps the great information coming!

Donation Information

I would like to make a donation in the amount of:

 $500 $200 $100 $50 $20 $10 $5 Other
Other:

I would like this donation to automatically repeat each month

Tribute Gift

Check here to donate in honor or memory of someone
Check here if this is a memorial gift
Name of person to be honored:
Send acknowledgement via email
Send acknowledgement via postal mail
Email Name:
Email:
Name:
Address:
City:
State :
Province:
Country:
Postal Code:

Donor Information

First Name:
Last Name:
Email:
Please do not display my name publicly. I would like to remain anonymous
Add me to your mailing list

Leave a Reply Cancel reply

Connect with:
Facebook

Your email address will not be published. Required fields are marked *

About the Author

Mike DeHaan

Mike DeHaan applies his Bachelor of Math in Computer Sciences degree, years of Cobol programming and quality assurance (including testing credit card interest calculations) to research and present mathematical theory for the … Read Full Profile

Follow Decoded Science

  • Facebook
  • Google+
  • Twitter
  • Pinterest
signupheredailydosedecsciv2


Science Everyone’s Talking About

  • What is a Metal? Chemistry, Metals, and the Periodic Table of the Elements What is a Metal? Chemistry, Metals, and the Periodic Table of the Elements Metals are often confusing. Let’s explore the definitio… under Ask the Expert
  • Climate Change Checkup, June 2017 Climate Change Checkup, June 2017 May was warm, but not a record. Decoded Science explore… under Headlines
  • Chemistry of the Giant Hogweed: A Macabre Cloud with a Silver Lining? Chemistry of the Giant Hogweed: A Macabre Cloud with a Silver Lining? The Giant Hogweed is a hot topic in the U.S., U.K. and… under Chemistry
  • As Condensation Droplets Form and Combine, What Forces are at Work? As Condensation Droplets Form and Combine, What Forces are at Work? Moisture condenses onto glass, forming condensation dro… under Ask the Expert
  • How Do Helicopters Fly? How Do Helicopters Fly? Helicopters fly by the same aerodynamic principles as a… under Aviation

Today’s Most Popular Science Articles

  • Red Ants Reaction and Treatment: What is a True Allergic Reaction to Red Ants? Red Ants Reaction and Treatment: What is a True Allergic Reaction to Red Ants?
  • Stomach Flu Cramps: Is There Anything You Can Do For Norovirus Symptoms? Stomach Flu Cramps: Is There Anything You Can Do For Norovirus Symptoms?
  • Which Chemical Bond is Stronger: Ionic vs. Covalent Bonds Which Chemical Bond is Stronger: Ionic vs. Covalent Bonds
  • Introducing the Factorial: the Exclamation Mark of Math Introducing the Factorial: the Exclamation Mark of Math
  • Introducing Math Symbols for Union and Intersection Introducing Math Symbols for Union and Intersection
  • Cross Multiply to Solve Equations with Fractions Cross Multiply to Solve Equations with Fractions
  • Practical Uses of Matrix Mathematics Practical Uses of Matrix Mathematics

© 2026 DecodedScience

MENU
  • Home
  • Headlines
  • General Science
  • Applied Science
    • Calculations
    • Economics
    • Engineering
      • Aviation
      • Civil Engineering
    • Medical Science
      • Health
      • Neuroscience
      • Oncology
      • Veterinary Science
    • Political Science
      • Polling
    • Mathematics
    • Technology
      • Artificial Intelligence
      • Computing
      • Electronics
      • Gadgets
    • Social Science
      • Cognitive Science
      • Psychology
      • Sociology
      • Anthropology
        • Linguistics
  • Physical Science
    • Archaeology
    • Astronomy
    • Chemistry
      • Materials Science
    • Geoscience
      • Climate Change
      • The Environment
      • Geology
      • Meteorology
      • Oceanography
    • Life Science
      • Biology
        • Botany
        • Zoology
          • Marine Biology
          • Entomology
          • Microbiology
        • Paleontology
        • Ecology
    • Nuclear Science
  • Theoretical Science
    • Physics
    • Math Theory
  • About Us
    • Contact Decoded Science
    • Ask the Expert
    • Meet Our Experts
    • Meet Our Sponsors:
    • Browse All Articles
    • Subscribe
    • Privacy Policy
    • Terms of Use Agreement
  • Support Decoded Science