In brief:
- Pillar A’s underlying question is who gains the most when the headcount goes down. The rules from Articles 5, 7 and 8 serve everyone, the leaders included. They fail only when the pay plan rewards breaking them.
- Meta raised 2026 capex guidance by $10B and planned ca. 8,000 cuts in the same quarter. It also granted its executives options that vest in full only if its market value reaches $9 trillion by 2031. The payout depends on market value alone, regardless of how the company gets there.
- The finance evidence is mixed. Buybacks frequently add value but, where rewards are tied to EPS, they also come with cuts in jobs and investment. Even staunch defenders of buybacks blame the problem on short-term pay design.
- In Germany and France, the law puts employees on boards, where pay and restructuring are decided.
- Pillar A includes 7 steps that any company can take. It starts with reading the pay plan.
Unlike my previous RETAIN articles, I don’t have a personal story that relates to leadership pay plans driving layoffs. This article has its source in a very different place.
For years, I have built checks and balances for processes that run at scale. The lesson I’ve learned is quite simple – a process that relies on everyone’s good intentions alone will inevitably fail. A process with the right checks and both short and long-term controls keeps working when put under extreme pressure.
People management has taught me the same thing. Even the best people may still respond in suboptimal ways to incorrectly designed incentives. So the question is not whether our people (or leaders) are good, but rather what behaviours the system rewards.
Pillar A: Aligned Leadership Incentives
I ended Articles 7 and 8 by stressing two rules. The company should publish the steps it will follow to improve its results. It should also keep its people regularly updated on every step. These rules serve leaders as much as everyone else. They fail only when the pay plan rewards breaking them or not caring at all. The essence of pillar A is mitigating the risk of that happening.
Pillar A. Executive pay and bonus structures should not be designed in a way that makes headcount cuts the fastest way to obtain personal payouts. It is essential to have well-aligned and prioritized incentives. They need to be tied to long-term performance, not only to quick wins. For example, a hasty big layoff will do miracles to EPS but it will also hurt the company in the medium term, when good people leave and systems and processes start falling apart.
To be clear, buybacks are not necessarily bad. Neither is every layoff. The finance evidence is mixed, and I’m covering it below. Several practices can turn a quick win into a personal payout: EPS-based bonuses, buybacks that lift EPS, cuts that translate into lower costs, and equity awarded based on market value alone. The question is who gains first when the headcount falls, and what checks stand in the way.
Pillar A works on two levels:
- Checks and balances. Someone is accountable, 12 months down the line, for the result of a layoff. The board tracks workforce signals, such as the exit of top performers. Employees have a voice where the pay is set.
- Incentives. The pay rewards lasting results, not just quick wins. No part of the plan ends with a higher reward simply because people have been cut.
One example of a quick win: EPS
Let’s take the following example. A buyback results in a lower share count, meaning the same profit is split over fewer shares. EPS goes up. A layoff cuts the costs, meaning the profit itself goes up. EPS goes up as well. Whether either of these raises the leaders’ payouts depends on their pay plan. So the plan is the first thing to scrutinize.
One of the best approaches I’ve found comes from Almeida, Fos and Kronlund, Journal of Financial Economics, 2016:
- They compared companies that would have just missed the analyst EPS forecast without a buyback with the companies that met it.
- The companies near the miss were more likely to buy back shares.
- Over the next four quarters, those companies cut employment, capex and R&D more than the companies that just met the forecast.
- The extra cuts averaged ca. 5% of headcount, 10% of capex and 3% of R&D.
- The effect was weaker or absent in companies that didn’t mention EPS in their proxy statements (the filings where a company explains executive pay to its shareholders).
The bottom line: where pay and targets are built around EPS, trading jobs for buybacks is a documented pattern.
What finance research says about EPS-linked pay
Sajjadiani, Kammeyer-Mueller and Benson surveyed 1,620 stores of a large US retailer over 22 months (Academy of Management JournalThis is not a clean story and both sides have their evidence.
For the critics:
- Almeida, Fos and Kronlund (above): EPS-driven buybacks were accompanied by cuts in employment, capex and R&D.
- The UK study below found that companies with EPS targets in pay contracts invested less than firms without them.
- Graham, Harvey and Rajgopal (survey of 401 US financial executives, Journal of Accounting and Economics, 2005): 78% said they would give up some economic value to report steady, predictable earnings every quarter. 55% said they would avoid starting a very valuable project if it meant missing the quarter’s consensus.
- Edmans, Fang and Lewellen (Review of Financial Studies, 2017): in the years when executives’ equity vests, R&D investment grows more slowly and companies are more likely to report earnings that land at, or just above, the analysts’ forecast.
For the defenders:
- Alex Edmans (London Business School) cites a study in which companies that buy back stock earned 12.1% more in stock returns than comparable firms over the next four years. That matches Ikenberry, Lakonishok and Vermaelen (1995), which covered buybacks announced in 1980-1990.
- He also cites evidence that buybacks happen when growth options are poor and cash is spare. In his view, firms usually decide on investment first, then buy back with what is left. That is an average, not a rule for every company.
- Almeida, Fos and Kronlund also found no loss of value on average from EPS-driven buybacks. The stock reaction around the earnings announcement was about the same as for firms that beat the forecast without one. That is, however, a very short window – it doesn’t show what the cuts cost the company in later years.
- A UK study for the government (Edmans and PwC, Sep’19, FTSE 350 companies, 10 years) found no sign that buybacks inflated executive pay or crowded out investment. No company in the sample used a buyback to reach an EPS target it would otherwise have missed.
Here is where both sides meet:
- Edmans places the root problem in short-term pay schemes, not in buybacks.
- The UK study calls the EPS-target link a correlation, not causation. Struggling companies may both cut investment and adopt EPS targets.
- Both kinds of study average across many companies. Neither tests what happens where a pay plan rewards a quick win and no checks stand in the way.
- In both cases the pay design is the issue.
Pillar A is not an anti-buyback manifesto. It’s meant to test if a leader’s payout may be achieved by a layoff itself, regardless of whatever happens to the workforce and the company.
Meta: more capex, fewer roles, new executive options
Here is a well known example of incentives gone wrong. Over the period of two months, Meta announced cuts of ca. 10% of its staff, raised its 2026 capex guidance by $10B and disclosed executive options that vest in full only at a $9 trillion market cap by 2031. The award depends on market value alone. This is a live test of pillar A’s question: who gains first when headcount falls?
- Apr’26: a cut of ca. 8,000 jobs is announced in an internal memo (ca. 10% of staff). A cancellation of ca. 6,000 open roles is also mentioned (as reported by TNW).
- Same month: Meta raises its 2026 capex guidance by $10B, to $125-$145B. Capex was $72.2B in 2025. Q1’26 revenue was $56.3B, up 33%.
- The headcount was 77,986 at the end of Mar’26. The CFO told the investors that the company would “reduce the size” of its employee base in May.
- The payroll was not simply shrinking. Total costs rose 35% to $33.4B in Q1. The CFO named infrastructure and employee compensation as the drivers, with compensation growth coming from technical hires, particularly AI talent.
- Late Mar’26: SEC filings disclosed option grants to six senior executives. The options would vest in full only if Meta’s market value grew from ca. $1.5 trillion to $9 trillion by 2031.
- Equilar estimated up to $625.6M in options per executive, and up to $921M for some once restricted stock was added (Fortune). Meta defends the plan as a big bet that pays only if the shareholders gain.
- A compensation adviser quoted by Fortune warned that awards like this “create undue risk-taking.”
Impact: the award depends on market value alone. Over the same two months, capex went up and people were cut. As part of Pillar A, I’m not saying that is unconditionally wrong, but I’m raising one question: does the payout depend on how well the changes are executed? A market-value measure by itself does not distinguish between a cut done well and a cut done badly. The target date is 2031, five years away. Rushed steps taken now could hurt growth over that whole period, even if they lift the market value at first. The real question to ask is which short-term controls balance decisions being made for a payout that far away. A second question: how do we stop executives from laying people off and disincentivizing the survivors, instead of motivating everyone to row harder in the same direction?
Oracle: allegations, kept apart from facts
Note: this story is not fully verified. None of the claims below have been tested or accepted by a court, and Oracle has not answered them.
What is reported as fact:
- Oracle cut ca. 21,000 positions in the fiscal year that ended in May’26, ca. 13% of its workforce. Another round began in Sep’26 (CTech, Sep 24th, 2026).
What is alleged:
- Apr’26: a former employee with a 30-year tenure wrote on LinkedIn that the cuts seemed to hit senior people with unvested stock. She said that she did not know how people were chosen. Others raised the same concern in online forums (People Matters, Apr 14th, 2026).
- The staff who were laid off lost unvested awards at once (New York Post, as cited by People Matters).
- Oct’26: three former employees filed a suit. They alleged that their terminations were timed just outside their RSU vesting dates, with 10 days’ notice, while others got more than two months. They put the loss at almost $1M in total (Redwood City Pulse, as excerpted by Autobound).
- Oracle has not said how it chose who would go.
Why am I mentioning this case here at all? It matters, because it illustrates an important test. Who decides a notice date? What does the company keep if unvested equity is forfeited? Who ultimately benefits? Any company can ask these questions as part of its own process.
Impact: if a notice date can change a person’s pay by a large amount, a published rule should set the date. Otherwise, whoever picks the date can gain from it. That is Article 7’s rule again, applied to equity.hese cases, employees could only take a guess at the order of steps. That is the gap T is meant to close.
Outside the US: who sits where pay is decided
The examples above were US-centric. Other countries have resolved issues related to pillar A in different ways.
In Germany, companies with more than 2,000 employees give half of the supervisory board seats to elected employee representatives. The chair, usually from the shareholder side, holds a casting vote (INSEAD, 2018).
Evidence (Kim, Maug and Schneider; 142 DAX and MDAX firms, 1990-2008):
- In industry downturns, companies with half-employee boards kept ca. 8% more employees. Their staff earned ca. 3.3% less on average. The authors interpreted the pay gap as the price of the insurance. The protection went to white-collar and skilled blue-collar workers. Unskilled blue-collar workers got no significant protection. Over the full cycle, the shareholders were no better or worse off.
Newer evidence:
- Jäger, Schoefer and Heining, Quarterly Journal of Economics, 2021; German firms around a 1994 law change: one-third worker seats on the board did not change the wages. If anything, they raised capital formation. There was no sign that the companies cut investment to pay for higher wages.
- In Dec’24, Volkswagen agreed with its works council and the union IG Metall to cut more than 35,000 German jobs by 2030, through early retirement and buyouts. The deal ruled out compulsory redundancies and plant closures until the end of 2030, and a planned pay cut of ca. 10% was dropped (Euronews).
In France, since the PACTE law of 2019, companies with more than 1,000 employees in France, or more than 5,000 worldwide, must seat at least two employee directors on a board of more than eight members. Boards of eight or fewer must seat one.
According to France Stratégie (Oct’23), 100% of eligible CAC 40 companies and 98% of eligible SBF 120 companies complied. In Jun’23, 82% of SBF 120 boards had a pay committee with at least one employee director, up from 55% in 2019. The research is divided on the results, but where an effect shows, it is generally positive. The studies cover four areas: profitability, innovation, sharing of value with employees and sustainable development.
The EU rules on pay policies prescribe that a listed company’s pay policy must explain how it considered the pay and employment conditions of its employees (Article 9a(6) of the Shareholder Rights Directive).
Here are some pay plan examples from Europe and North America:
- WTW (1,070 public-company disclosures, 18 markets, fiscal years ending May’24 to May’25): 71% of North American companies and 81% of European companies put at least one people-related metric in executive incentive plans. The usual metrics are engagement, succession, culture and retention.
- WTW also reports that only 9% of the ESG metrics used by S&P 500 companies are connected to long-term plans.
What A prescribes
- Read the pay plan first. List every metric in the executive bonus and equity plans. Mark each one that rises when headcount falls: EPS, operating income, margin, share price. Then ask which one would also rise if the steps from Articles 5, 7 and 8 were skipped.
- Extend the outcome check to cuts. In Article 5, the fifth part of the headcount gate is a 12-month outcome check, tied to the pay of the people who requested and signed a role. Apply the same idea to cuts: whoever signs a cut is accountable for its 12-month outcome. At 12 months, someone compares what was promised when the cut was made with what has been delivered. Examples of measures: the savings actually achieved, the exit of top performers, roles rehired or refilled by contractors, and service or delivery problems.
- Put a workforce check on the board’s own scorecard. Two proposals: voluntary attrition among top performers in the 12 months after a cut, and the share of stayers who say they have considered leaving. Define the measures before the layoff, not after it. The aim is to ensure that no scorecard component pays more when the steps covered in my previous articles are skipped.
- Explain the same-cycle case. When a buyback or a capex increase lands in the same earnings cycle as a layoff, the board explains in writing and without corporate newspeak why the cut was needed even without that spend. Otherwise the staff and the investors will be left guessing whether it was the people who paid for the buyback.
- Neutralize buybacks in EPS targets. Adjust EPS-based targets for the effect of buybacks, or use measures that buybacks are unlikely to move. Examples: operating income, revenue growth, or EPS calculated on the share count before the buyback.
- Give the workforce a voice where the pay is set. France does it by law for large companies. Elsewhere, a standing consultation with the pay committee could be a good start. Say in the pay policy how workforce pay was considered.
- Publish the rule for notice dates and unvested equity. This is Article 7’s rule, applied to equity. No one should lose their hard-earned rewards because notice dates get manipulated.
What aligned incentives don’t guarantee
- A good pay plan does not translate into good decisions. Leaders can still make bad calls on who gets cut and when.
- A pay plan cannot make a leader care. It can only make sure that not caring has no positive financial impact for that leader.
- Some cuts may be right. Some buybacks add value. Some quick wins are real wins. What matters is that the pay plan and the board can tell them apart.
- Measures can be gamed. An engagement score can be raised without helping anyone, and a 12-month review can be softened after the fact. That is why each measure is defined before the cut, as step 3 says.
Next in this series
I: institutional knowledge as an asset. When the cuts are done, what actually walks out the door?
Sources
Meta
- Meta, Q1 2026 results, Apr 29th, 2026
- Yahoo Finance, Apr 29th, 2026
- The Next Web, Apr 23rd, 2026
- Fortune, Mar 28th, 2026
Oracle (allegations)
- CTech, Sep 24th, 2026
- People Matters, Apr 14th, 2026
- Redwood City Pulse, Oct 23rd, 2026, as excerpted by Autobound
Research
- Almeida, Fos and Kronlund, The Real Effects of Share Repurchases, summary by Kronlund, Harvard Law Forum, Feb 8th, 2016
- Edmans, Do Share Buybacks Really Destroy Long-Term Value?, Harvard Law Forum, 2020
- Edmans and PwC for BEIS, London Business School, Sep’19
- Kim, Maug and Schneider, Labor Representation in Governance as an Insurance Mechanism, ECGI working paper 411/2014
- Ikenberry, Lakonishok and Vermaelen, Market Underreaction to Open Market Share Repurchases, NBER working paper 4965
- Graham, Harvey and Rajgopal, The Economic Implications of Corporate Financial Reporting, NBER working paper 10550
- Edmans, Fang and Lewellen, Equity Vesting and Investment, London Business School research page
- Jäger, Schoefer and Heining, Labor in the Boardroom, IWH summary
Law and pay practice
- INSEAD Knowledge, Giving workers equal representation on the board, 2018
- White & Case, PACTE law alert, May’19
- France Stratégie, fourth PACTE report, Oct’23 (full report on Vie publique)
- Shareholder Rights Directive, Article 9a, on legislation.gov.uk
- WTW, human capital in executive incentive plans, Jan 22nd, 2026
- Euronews, Volkswagen agreement with the union, Dec 20th, 2024


