countries forming coalitions to fund a public good
Your Holding Company owns five PE portfolio companies. All five would benefit from a shared compliance platform - estimated $2M to build, saving each company $800K/year. Total value: $4M/year on a $2M Capital Investment. But when you ask each CEO to chip in $400K, three say 'we'll wait and see if the others build it first.' The platform never gets built. Welcome to the public good problem.
A public good is something where one party's use doesn't reduce its value for others, and you can't practically stop non-payers from benefiting. This makes rational actors underpay - even when the collective ROI is enormous - because each party's Dominant Strategy is to let someone else fund it. Operators solve this through consortium structures, cost sharing rules, or central Allocation from a Holding Company.
A public good has two properties that make it different from normal Capital Investments:
The combination is what makes funding break down. Any single good that's easy to share and hard to restrict access to is a public good - industry standards, shared threat intelligence, common data infrastructure, even a clean break room.
The node description frames this as countries forming coalitions, but the mechanics are identical when business units, PE portfolio companies, or vendors need to jointly fund something that benefits everyone.
Public goods show up constantly in Operations and they quietly destroy value when mishandled:
The P&L impact is indirect but large. When public goods go unfunded, each team builds their own inferior version. You pay 5x the Implementation Cost across the portfolio for 0.5x the quality. The surplus that a shared investment would create simply evaporates.
Consider three business units that would each get $500K/year in value from a shared analytics platform costing $600K to build.
This is the public good trap. Individual rationality produces collective stupidity. Game Theory predicts this: when your Outside Option is 'benefit without paying,' contributing is never your best move unless the structure forces it.
When one company builds the platform, it generates a positive Externality - value that spills over to others who didn't pay. The Externality framework from your prerequisites tells you to measure this: what does each party's participation add to (or subtract from) others' outcomes? That measurement is the foundation for fair cost sharing.
Solving public goods means building a consortium where enough parties commit to make the investment viable. Three mechanisms work:
Imagine each party privately decides how much to contribute. Each party's Utility Maximization says: contribute the minimum possible, because every dollar someone else contributes gives me the same benefit. The equilibrium of voluntary contributions is always below the Efficient Allocation. This isn't a behavioral flaw - it's the mathematically predicted outcome when rational actors face a public good.
Recognize a public good situation when all three conditions hold:
When all three hold, don't rely on voluntary funding. Choose your mechanism:
A Holding Company has four PE portfolio companies (A, B, C, D). A shared customer data platform costs $1.2M to build. Annual value to each company if all four participate: A gets $600K, B gets $400K, C gets $300K, D gets $200K. Total annual value: $1.5M. Each company's Outside Option is building their own version for $500K (inferior, covering only their own data).
Check if it's a public good. Multiple beneficiaries? Yes (four companies). Hard to exclude? Yes - once the platform exists, any portfolio company can query it. Non-diminishing? Yes - Company A running queries doesn't reduce Company B's ability to. All three conditions met.
Calculate collective surplus. Total value = $1.5M/year. Cost = $1.2M one-time. Even in year one, the surplus is $300K ($1.5M - $1.2M). Over a 3-year Time Horizon, that's $1.2M x 1 cost vs. $1.5M x 3 = $4.5M value. Massive ROI - but only if it gets funded.
Apply Shapley value for cost sharing. We compute each company's marginal contribution across all possible coalitions. Simplified result (proportional to value received): A pays $480K (40%), B pays $320K (27%), C pays $240K (20%), D pays $160K (13%).
Verify each party beats their Outside Option. Company A: pays $480K vs. $500K alone - saves $20K AND gets a better platform. Company D: pays $160K vs. $500K alone - saves $340K. Every party is strictly better off in the coalition.
Structure as binding agreement. The Holding Company executes a cost sharing agreement before development starts. Each company's Budget absorbs their share as overhead. No waiting, no hoping others pay.
Insight: The Shapley value allocation made every company better off than going alone. Without the forced structure, the likely outcome is Company A builds it alone ($1.2M for $600K value - negative first-year ROI) or nobody builds it. The Holding Company's authority to enforce cost sharing is what unlocks $4.5M in value over three years.
Five mid-size e-commerce companies each lose $2M/year to fraud. A shared fraud detection network (pooling transaction signals) would reduce losses by 60% - saving $1.2M/year per company, $6M/year total. The platform costs $3M to build and $500K/year to operate. No single company has authority over the others.
Identify the public good dynamics. Once the fraud network exists, any member's transactions make the model better for everyone (positive Externality). You can technically exclude non-members from API access, but the network effect means more members = better detection for all. This is a partial public good - excludable but with strong spillovers.
Model the incentive problem. Each company thinks: 'If I wait, the other four might build it. I'd then negotiate access later at a lower price.' If all five think this way, nobody moves. Even though the collective NPV is enormous ($6M/year savings - $500K/year cost on $3M build = positive by year one), individual Dominant Strategy is to wait.
Design the consortium structure. Equal split: $600K build cost each, $100K/year ops each. Each member's annual ROI: $1.2M savings - $100K ops cost = $1.1M net, on a $600K initial investment. Payback Period under 7 months.
Enforce with binding agreements. A consortium agreement with binding agreements specifying: minimum 3 of 5 must commit for the build to proceed. Data contribution requirements. Exit penalties equal to 2 years of ops cost ($200K). This removes the 'wait and see' option.
Handle late joiners. Companies that join after launch pay a 1.5x premium on the build share ($900K instead of $600K). This creates an incentive to join early - your Outside Option gets worse the longer you wait.
Insight: Without excludability or penalties, the fraud network never gets built despite a 7-month Payback Period. The consortium structure with binding agreements and late-joiner penalties converts a public good problem into a straightforward cost sharing problem. The key insight: someone has to design the rules before the good exists.
A public good creates value that you can't restrict and that doesn't diminish with use - which means rational actors underfund it even when collective ROI is enormous. The Dominant Strategy for each party is to let others pay.
Solve it with structure, not persuasion. Central authority (Holding Company mandates), fair arithmetic (Shapley value cost sharing), or binding agreements with penalties for non-participation. Voluntary goodwill consistently fails because the math works against it.
The operator's job is recognizing when a shared investment is a public good - then choosing the right funding mechanism before the value evaporates. Every month without the shared platform is surplus that never materializes.
Assuming good faith solves the problem. You email all five business unit heads: 'This platform helps everyone, let's all chip in.' Three agree in principle, none put it in their Budget, and the project dies in quarterly planning. The issue isn't bad intent - it's that each party's Dominant Strategy is to wait. You need a mechanism (authority, contracts, or penalties), not an appeal.
Treating a shared resource with capacity limits as a public good. A shared data analyst serves five teams. When Team A monopolizes their time, Team B gets nothing. That's a common resource with rivalry - solve it with resource allocation and prioritization, not public good funding mechanisms. Misdiagnosis leads to the wrong fix.
Your company has three product lines (X, Y, Z) that would all benefit from a shared ML feature store. Build cost: $900K. Annual value: X gets $400K, Y gets $350K, Z gets $250K. Each product line's Outside Option is building a local version for $450K. Calculate the Shapley value cost allocation and verify every product line is better off joining than going alone.
Hint: For a simple proportional Shapley allocation, weight the $900K build cost by each party's share of total value ($1M). Then compare each party's allocated cost to their $450K Outside Option.
Total value = $400K + $350K + $250K = $1M. Proportional shares: X = 40%, Y = 35%, Z = 25%. Cost allocation: X pays $360K (40% of $900K), Y pays $315K (35%), Z pays $225K (25%). Verification: X pays $360K vs. $450K alone (saves $90K, gets better platform). Y pays $315K vs. $450K (saves $135K). Z pays $225K vs. $450K (saves $225K). All three beat their Outside Option. The coalition is stable - no party wants to leave.
You're one of six equal-sized SaaS companies. An industry group proposes a shared API standard that would cost $1.8M to develop and would increase every company's Expansion Revenue by $500K/year by reducing integration friction for customers. You have no authority over the other five companies. Design a funding structure that prevents the 'wait and let others pay' problem.
Hint: Think about what makes waiting attractive (free benefit) and design mechanisms that either remove the free benefit or make early commitment cheaper than late joining. Consider minimum commitment thresholds, excludability through licensing, and late-joiner premiums.
Three-part structure: (1) Minimum viable coalition: set a threshold of 4 of 6 companies committing $300K each ($1.2M) before development begins. Remaining $600K covered by the API standard's licensing to non-members. (2) Excludability through licensing: the standard is open, but the reference implementation, testing tools, and certification badge require membership. Customers trust certified integrations - creating a Competitive Advantage for members. (3) Late-joiner penalty: companies joining after launch pay $450K (1.5x the founding member price). This makes early commitment the Dominant Strategy - wait and you pay more for the same thing. Expected Payoff for a founding member: $500K/year Expansion Revenue on a $300K one-time cost. Payback Period: under 8 months.
Public goods sit at the intersection of your three prerequisites. Game Theory explains why rational actors underfund them - when your Dominant Strategy is to let others pay, collective action fails without enforcement. Externality pricing measures the spillover value each participant creates for others, which is the foundation for deciding who should pay what. And cost sharing via the Shapley value provides the arithmetic to allocate costs fairly once you've formed the coalition. Downstream, public goods connect to Efficient Allocation - the operator's job is ensuring that shared investments actually get made when their collective ROI justifies them, rather than letting individual P&L optimization kill projects that create massive surplus. The concept also feeds into overhead management and consortium strategy in PE portfolio companies, where the Holding Company must decide which investments to centralize (fund as public goods) versus which to leave to individual brands.
Disclaimer: This content is for educational and informational purposes only and does not constitute financial, investment, tax, or legal advice. It is not a recommendation to buy, sell, or hold any security or financial product. You should consult a qualified financial advisor, tax professional, or attorney before making financial decisions. Past performance is not indicative of future results. The author is not a registered investment advisor, broker-dealer, or financial planner.