Intergenerational geometric transfers of income

This paper characterizes a family of geometric rules for intergenerational income transfers in an infinite stream by deriving them from axioms of consistency, continuity, independence, feasibility, and scale invariance.

The Gist

Imagine a giant, endless relay race that has been running forever and will continue forever. There is no starting line in the past, and there is no finish line in the future. The runners are generations of people.

In this race, every runner (generation) receives a baton made of money (income). The big question this paper asks is: How should we pass the baton?

Should the current runner keep the whole baton and run alone? Should they give it all to the next runner? Or should they split it? And what about the runners who came before us? Did they pass anything down?

This paper, written by four economists, proposes a very specific, mathematically elegant way to handle this "baton passing" called Geometric Rules. Here is the breakdown in simple terms.

1. The Setup: An Infinite Chain

Most people think of time as starting at birth (Generation 0) and moving forward. But this paper imagines time as an infinite line: ... -3, -2, -1, 0, 1, 2, 3 ...

• Negative numbers are our ancestors (the past).
• Zero is us (the present).
• Positive numbers are our descendants (the future).

The goal is to create a rule that takes the money everyone has and decides how much each generation keeps and how much they pass to the next person.

2. The "Geometric" Solution: The Leaky Bucket

The authors found that the best rules follow a pattern they call Geometric Rules.

Imagine every generation has a bucket of water (income).

• When a generation receives water, they keep a specific percentage (let's say 30%) for themselves.
• They pass the remaining 70% to the next generation.
• The next generation keeps 30% of what they received and passes 70% to the one after that.

This creates a "geometric" flow. The money trickles down the line, getting smaller and smaller with every step, but never quite disappearing completely.

Why is this special?
The authors prove that if you want a system that is fair, consistent, and doesn't break the rules of math, this "leaky bucket" method is the only way to do it.

3. The Five Golden Rules (The Axioms)

To find this solution, the authors set up five "laws of physics" for their money-passing system. If a rule breaks any of these, it's not a good rule.

1. Feasibility (Don't Create Money): You can't pass out more money than exists. You can't magically create wealth; you can only move it around.
2. Scale Invariance (The Dollar vs. The Cent): It shouldn't matter if we measure money in dollars or cents. If everyone's income doubles, the rule should just double the result. The proportions matter, not the units.
3. Independence of the Future (Don't Look Backward): How we treat the past shouldn't change just because the future changes. If the 100th generation gets a huge inheritance, it shouldn't change how much the 1st generation gets.
4. Consistency (The "What If" Test): This is the most important one. Imagine the race stops for a moment. The past runners have already left with their share. The current runner gets to keep their original money plus whatever was passed down to them from the past. The rule says: "If we re-calculate the distribution starting from now with this new total, the current and future runners should get the exact same amount as before." It ensures the system is stable and doesn't change its mind based on history.
5. Continuity (No Sudden Jumps): If the income stream changes just a tiny bit (a small breeze), the result shouldn't change drastically (a hurricane). Small changes in input must lead to small changes in output.

4. The Surprising Twist: How You Measure "Small Changes"

Here is where the paper gets really clever. The authors realized that "Continuity" depends on how you measure a "small change."

• The Taxicab Norm (The Standard Way): Imagine measuring the total distance of a change. If the total amount of money shifted is small, the rule is continuous. Result: The "Geometric Rules" work perfectly here.
• The Sup-Norm (The "Biggest Single Change" Way): Imagine you only care if any single person's income changes drastically. If you use this strict measure, the standard Geometric Rules break down. You have to find a special subset of rules where the "leakage" is controlled so no single generation gets overwhelmed.
• The Point-Wise Norm (The "Individual" Way): Imagine you check each person one by one. If you use this, the only rules that work are the extreme ones: either everyone passes everything (Full Transfer) or no one passes anything (No Transfer).

5. Why Does This Matter?

In the real world, we often argue about fairness between generations (e.g., climate change, national debt, pension funds).

• The Problem: We can't easily compare the happiness of a person in 1800 to a person in 2100.
• The Solution: Instead of guessing "happiness," this paper focuses on resources (money/income).
• The Takeaway: If we want a system that is fair, stable, and mathematically sound, we should adopt a Geometric Rule. This means every generation keeps a steady percentage of what they receive and passes the rest down. It creates a sustainable, flowing river of wealth that connects the infinite past to the infinite future without running dry or overflowing.

Summary Analogy

Think of the economy as a giant, infinite family dinner.

• The Geometric Rule is like a polite host who says: "I will eat 30% of the food passed to me, and I will pass the remaining 70% to the person on my right."
• This ensures that everyone gets a bite, the food never runs out (because it keeps circulating), and no one person gets greedy enough to eat it all.
• The paper proves that this specific "30/70" style of sharing is the only way to keep the dinner fair, consistent, and stable, provided you measure "fairness" in the right way.

Technical Summary

Here is a detailed technical summary of the paper "Intergenerational geometric transfers of income" by Algaba, Moreno-Ternero, Rémila, and Solal.

1. Problem Statement

The paper addresses the problem of intergenerational income transfers within an infinite, countable stream of generations. Unlike traditional intergenerational equity literature that focuses on ranking infinite utility streams (welfare), this study adopts a resourcist framework. The objective is to design allocation rules that map an initial infinite stream of income ($r$) to a new stream ($\phi(r)$), effectively modeling how income is transferred from past and present generations to future ones.

Key features of the model include:

• Infinite Domain: Generations are indexed by the set of integers $\mathbb{Z}$ (both past and future are unbounded), rather than the non-negative integers $\mathbb{N}$.
• Transfer Mechanism: The rule must determine how much income each generation retains and how much is passed to the next, creating a flow of resources across time.
• Goal: To characterize a family of rules that satisfy specific normative principles (axioms) regarding feasibility, fairness, and stability.

2. Methodology

The authors employ an axiomatic approach (social choice theory) to derive the allocation rules.

• Model Setup:
  • Let $L^1_+$ be the set of income streams $r: \mathbb{Z} \to \mathbb{R}_+$ such that $\sum |r_i| < \infty$.
  • An allocation rule is a function $\phi: L^1_+ \to L^1_+$.
• Candidate Family: The authors propose a family of Geometric Rules ($\mathcal{G}_\lambda$), parameterized by a sequence $\lambda: \mathbb{Z} \to [0, 1]$.
  • In a geometric rule, generation $i$ retains a share $\lambda_i$ of the income it receives (from its own endowment plus transfers from the past) and transfers the remaining share $(1-\lambda_i)$ to generation $i+1$.
  • The formula for the allocation to generation $i$ is:
    $$\phi^\lambda_i(r) = \lambda_i \sum_{j \leq i} \left( \prod_{k=j}^{i-1} (1-\lambda_k) \right) r_j$$
• Axioms: The study tests allocation rules against the following axioms:
  1. Feasibility: Total allocated income $\leq$ Total initial income.
  2. Scale Invariance: The rule is independent of the currency unit ($\phi(\alpha r) = \alpha \phi(r)$).
  3. Independence of a Future Income: Changes in the income of generation $j$ do not affect the allocation of any generation $i < j$.
  4. Consistency: If the allocation is re-evaluated at generation $j$ (assuming past generations have left with their awards and the residual income is added to $j$'s endowment), the allocation for $j$ and all future generations remains unchanged.
  5. Continuity: Small changes in the input stream result in small changes in the output stream. The paper primarily uses the $L^1$ (taxicab) norm but also explores sup-norm and point-wise convergence.

3. Key Contributions and Results

A. Main Characterization Theorem (Theorem 1)

The central result establishes that the family of Geometric Rules is the unique family of allocation rules satisfying:

• Feasibility
• Scale Invariance
• Independence of a Future Income
• Consistency
• Continuity (in the $L^1$ norm)

Proof Logic: The proof constructs the parameter $\lambda_i$ directly from the rule's behavior on unit streams ($e_i$) and uses the axioms (specifically consistency and independence) to show that the rule must follow the recursive geometric structure for any arbitrary stream.

B. Uniform Geometric Rules

By adding Translation Invariance (shifting the time index does not change the allocation structure) to the main axioms, the authors characterize the Uniform Geometric Rules ($\mathcal{U}_\lambda$), where the retention rate $\lambda$ is constant across all generations ($\lambda_i = \lambda$).

C. The Role of Continuity Norms

A significant contribution is the analysis of how different continuity definitions alter the set of admissible rules:

1. $L^1$ Continuity (Taxicab): Characterizes the full family $\mathcal{G}_\lambda$.
2. Sup-Norm Continuity: Characterizes a sub-family $\mathcal{T}_\lambda \subset \mathcal{G}_\lambda$. These are rules where the cumulative transfer coefficients are bounded. This excludes rules where the "weight" of past income on current allocation grows unboundedly.
3. Point-wise Continuity: Characterizes a sub-family $\mathcal{P}_\lambda \cup \{\phi_0\}$. These rules require that for every generation $i$, there exists a past generation $j < i$ with a retention rate $\lambda_j = 1$ (effectively blocking infinite chains of transfers).
   • Implication: Under point-wise continuity, the only Uniform rules are the extreme cases: No-Transfer ($\lambda=1$) and Full-Transfer ($\lambda=0$).

D. Balance and Feasibility

• Feasibility: All geometric rules satisfy feasibility.
• Balance (No Waste): A geometric rule satisfies balance (Total Allocated = Total Initial) if and only if it belongs to the sub-family $\mathcal{B}_\lambda$, defined by the condition $\prod_{k=i}^{\infty} (1-\lambda_k) = 0$. This ensures that income is not "lost" in an infinite chain of transfers without ever being consumed.

E. Idempotency

Adding the Idempotency axiom ($\phi(\phi(r)) = \phi(r)$) restricts the solution to only the two extreme uniform rules:

• No-Transfer Rule: Each generation keeps everything ($\lambda=1$).
• Full-Transfer Rule: Each generation passes everything to the next ($\lambda=0$), though in the infinite past/future context, this results in zero allocation to any specific generation unless balanced carefully.

4. Significance and Discussion

• Resolution of Impossibility Results: Unlike the literature on intergenerational welfare (e.g., Diamond, 1965), which often yields impossibility results when combining impartiality and continuity, this resourcist framework successfully characterizes a meaningful, constructive family of rules.
• Infinite Past and Future: The use of $\mathbb{Z}$ (unbounded past and future) is crucial. It allows for a distinct analysis of "backward" conditions (like point-wise continuity requiring a past blockage) versus "forward" conditions (like balance requiring infinite dissipation of transfers).
• Normative Implications: The paper demonstrates that the choice of the continuity norm is not merely technical but has profound normative consequences.
  • $L^1$ continuity allows for flexible, long-term transfer chains.
  • Point-wise continuity forces a "reset" mechanism (a generation that keeps everything) to ensure stability, effectively limiting the reach of intergenerational transfers.
• Connection to Existing Literature: The results extend finite-population fair allocation models (e.g., claims problems, river sharing) to the infinite horizon, highlighting how continuity axioms act as the binding constraint in infinite settings.

Conclusion

The paper provides a rigorous axiomatic foundation for intergenerational income transfers. It identifies Geometric Rules as the natural solution for infinite streams, showing that the specific properties of these rules (uniformity, balance, or extreme behavior) depend critically on the chosen continuity norm and additional stability axioms like translation invariance or idempotency. This offers a flexible toolkit for policymakers or theorists modeling resource distribution across unbounded time horizons.