Compensation and Incentive Design

Core Concepts

Pay transparency

Transparency is not a single policy — it exists on a spectrum from disclosing salary ranges in job postings to publishing every employee's individual salary publicly. The level of disclosure changes what problems it solves and what problems it creates.

What transparency does well. When salary ranges and the criteria behind pay decisions are disclosed, the likelihood that gender or racial bias silently shapes compensation drops. The mechanism is straightforward: visibility enables comparison, and comparison creates accountability. Studies of Canadian university faculty after mandatory salary disclosure found gender pay gaps reduced by 20–40%. Danish pay gap reporting requirements produced approximately a 2 percentage point reduction from a 17% baseline. In US academic settings, pay transparency reduced pay variance within departments by nearly 20%. (NBER Working Paper on Pay Transparency and the Gender Gap; OECD: Pay Transparency Tools)

What transparency can backfire on. Making pay visible does not automatically make it fair — it makes existing disparities vivid and actionable. Equity theory operates on perceived ratios of contribution to reward. When those ratios were opaque, workers estimated them through noisy signals like job level and tenure. Transparency converts estimation into knowledge. If the underlying structure was already unequal, transparency amplifies the fairness violation rather than resolving it. An engineer who discovers a 30% equity grant gap versus a peer — from a transparent pay band rather than a rumor — experiences the disparity as concrete and grievable. The same economic fact, invisible under opacity, becomes an organizational event once disclosed. (PAVE: Gender Pay Inequity in Software Engineering)

This means transparency is only safe to deploy after you have audited and corrected the underlying structure. Disclosing a broken pay system accelerates its consequences.

Formula-based pay

One response to the bias-through-discretion problem is to remove discretion entirely: replace manager judgment in pay-setting with a published formula applied consistently to everyone.

Buffer and GitLab are the most prominent implementations. Buffer's formula multiplies a role benchmark by a cost-of-living factor and an experience multiplier. GitLab's is more granular: SF benchmark × location factor × level factor × experience factor × contract factor × exchange rate. Both companies publish the full formulas and individual salaries openly. Buffer reports that the open pay system builds trust while maintaining accountability. Both trade manager discretion for auditability.

The bias-reduction logic is clean: if no human is making pay decisions, no human bias can enter pay decisions. Buffer explicitly recognized that its earlier discretionary framework left room for bias, and moved to formulas specifically to close that vector.

Tournament theory and prize gaps

Tournament theory (Lazear & Rosen, 1981) explains why organizations maintain steep pay gaps between adjacent hierarchical levels, particularly at senior ranks. The key insight is that it is the size of the prize gap — the difference in pay between one level and the next — rather than absolute compensation amounts, that drives effort. Workers compete not for their current salary but for the salary of the level above them.

In engineering orgs, this manifests as intentionally wide gaps between senior engineer and staff engineer, or between staff and principal. A Danish study of 2,421 companies found strong empirical support for the relationship between tournament structure and firm performance. (Journal of Labor Economics, Vol 17 No 2)

Tournament theory is a description of a mechanism, not a prescription. The same mechanism that motivates individuals to compete for promotion prizes also suppresses the cooperative behaviors — knowledge sharing, mentoring, unowned problem-solving — that high-performing engineering organizations depend on.

What this means for org design: large prize gaps between levels are not free. They buy individual effort at the cost of team collaboration. In roles where output is primarily individual and measurable (e.g., sales), tournament structures are well-matched. In engineering, where much of the high-leverage work is cross-functional, invisible, and relational, the same structure creates perverse incentives. Bonus pools that create explicit competition between peers amplify this problem.

Equity and RSU structures

RSUs (Restricted Stock Units) are the dominant equity vehicle in public tech companies, and a major component of total compensation — often 30–50% of target total comp at senior levels.

Cliff vesting and retention. The industry standard is a 4-year vesting schedule with a 1-year cliff: no equity vests until the first anniversary of the grant, after which vesting proceeds monthly or quarterly. 65% of workers report that increased equity would extend their tenure; 84% express interest in equity as a motivating mechanism; 48% consider it the most effective engagement mechanism (Morgan Stanley State of Workplace 2025). The cliff is a deliberate retention device: it ensures the employee has contributed long enough to be worth the organizational investment.

Price decline risk. RSUs are often pitched as "guaranteed value" — unlike options, they don't expire worthless. But the realized value of an RSU grant depends entirely on the stock price at vesting, which can diverge sharply from the price at grant. Amazon's stock declined approximately 50% from mid-2021 to late 2022, causing employee compensation to fall 15–50% below target levels and the equity portion to crash by 50%. Amazon responded by granting a record 138 million RSUs in Q2 2022 and raising its salary cap to $350,000. Meta's total compensation showed the reverse: 30% potential increases in RSU-based compensation as stock price recovered in 2024. (FAANGFire)

This volatility creates a structural problem: total compensation promises in offer letters are hypothetical, and the employee bears stock price risk with no corresponding control. At any given moment, a portion of your engineering workforce may be operating on compensation that is materially below what they were recruited on — a flight risk you will not see until attrition picks up.

Location-based pay

Approximately 71% of companies employ location-based pay adjustments for remote workers as of 2025–2026, creating systematic pay differentials for identical roles and levels across geographic regions. (Second Talent: Remote Work & Hiring Statistics 2026) This is the dominant model for distributed engineering teams.

The logic is cost-of-living alignment: role benchmark × location factor, as in both Buffer and GitLab's formulas. Buffer uses three geographic bands: 1x (high COL), 0.85x (average), 0.75x (low). GitLab's location factor is more granular.

Location-based pay is operationally straightforward but creates equity tensions. Eastern European skilled professionals earn 40–60% less than Western European equivalents for comparable technical work. Remote workers on average earn 4–7% more than office counterparts, but location-based adjustments can reverse this advantage for workers in lower-cost regions. Some organizations are shifting toward job-based pay (uniform pay for a role regardless of location) to remain competitive in tight talent markets.

EU Pay Transparency Directive

The EU Pay Transparency Directive (Directive 2023/970) was adopted in May 2023. EU member states must transpose it into national law by 7 June 2026 — a deadline the European Commission has confirmed it will not extend, even as most member states have not yet published draft legislation. (EUR-Lex: Directive 2023/970; Ogletree: EU Pay Transparency Directive Progress)

What it requires of employers:

"Equal value" is broader than "same title." The Directive defines work of equal value using four objective, gender-neutral criteria: skills, effort, responsibility, and working conditions. Relevant soft skills must not be undervalued. Roles contributing comparably to organizational success can be in scope even with different titles. (EUR-Lex: Equal Pay Summary)

Key Principles

1. Transparency is safe only after structural correction. Deploy pay transparency after auditing and fixing underlying gaps, not as a mechanism for detecting them. Disclosure of a broken pay structure accelerates its organizational consequences — it turns a latent problem into a visible crisis.

2. Formula design is a moral choice, not a technical one. Formula-based pay removes individual discretion but encodes organizational values into the formula's factor choices. Auditing the formula for built-in biases is as important as having a formula.

3. Prize gaps have collaborative costs. Tournament-style pay dispersion between levels buys individual effort and sacrifices team cooperation. Calibrate level gaps with an eye on the work your organization actually needs done — not just on motivating promotion-seeking behavior.

4. RSU-based total compensation is a target, not a promise. Any engineering organization where senior-level total compensation is RSU-heavy is implicitly running a compensation program that varies with market conditions. Treat stock price declines as a compensation risk, not just an accounting event.

5. Location-based pay is a formula choice with equity implications. Geographic pay bands are a legitimate cost management tool, but they create differentials that interact with EU "equal value" analysis. If the same role at different locations is paid differently, that difference must be justifiable by objective criteria.

6. The EU Directive creates structural, not procedural, requirements. Posting a salary range in a job ad is the easy part. The hard part is building the job evaluation framework that assigns roles to pay bands on objective criteria, maintaining pay gap data, and having a remediation process ready to trigger at 5%.

Worked Example

Scenario: A 600-person engineering organization, distributed across the EU and US, uses discretionary pay-setting with manager approval. It is Q4 2025. The Head of Engineering is preparing for EU Directive compliance.

Step 1 — Audit the existing structure. Commission a regression pay equity analysis grouped by role level and country. This identifies whether gender or other protected characteristics explain residual variation after accounting for level, tenure, and location. If gaps exist, you need to close them before the Directive's reporting begins, because you will not want your first published report to reveal a 5%+ unexplained gap.

Step 2 — Build a job evaluation framework. The Directive requires pay comparison across roles of "equal value" — defined by skills, effort, responsibility, and working conditions. Map your engineering roles onto these four criteria in writing. This is the evidence base for defending pay differences between roles (not just within them) to a regulator or in litigation.

Step 3 — Define pay bands with objective, gender-neutral criteria. Each pay band needs a documented rationale. If you are using location factors, document why and how they are set. Formula-based pay (Buffer, GitLab) makes this natural; discretionary systems must retrofit the documentation onto decisions that were never recorded.

Step 4 — Update recruitment process. Remove salary history from application forms and interview scripts immediately — this is a violation risk regardless of whether your member state has completed transposition. Prepare salary range disclosure materials for every open role. The range must be based on the objective criteria in Step 3, not on candidate negotiation posture.

Step 5 — Establish the 5% monitoring cadence. Build a quarterly compensation review that flags any pay category where the gender pay gap exceeds 5% and cannot be justified by objective criteria. You have six months to remedy before a joint pay assessment is triggered. The monitoring process should be owned by People/HR with data infrastructure, not by individual managers.

Step 6 — Prepare the joint assessment protocol. Before you need it, document how a joint pay assessment would run: who the worker representatives are, what data they would receive, what remediation measures are in scope, and how the completed assessment would be published to employees. Having this in a drawer before you need it is far cheaper than constructing it under a six-month deadline.

Compare & Contrast

Formula-based pay vs. band-based discretionary pay

The key trade-off: formula-based pay makes the organization's values explicit and auditable, but the formula cannot flex for individual context. Band-based pay preserves flexibility, but that flexibility is also where bias enters.

Cliff vesting vs. milestone vesting

Boundary Conditions

Pay transparency does not close gaps in negotiation. Transparency policies that post salary ranges still leave room for negotiation, and negotiation gender gaps persist. Research shows that making wage negotiability explicit in job postings narrows but does not eliminate gender gaps in negotiation behavior. Posting ranges without formula grounding still leaves the gap open.

Tournament theory works worst in cross-functional, invisible work. The prize gap mechanism was empirically tested in contexts where individual output is measurable — executive compensation, sales. In engineering, where high-leverage work is often collaborative and invisible (architecture decisions, incident prevention, knowledge transfer), steep level gaps can actively suppress the behaviors you most need. Principal-level pay structures need to be designed with this in mind.

Formula-based pay has geographic limits. Formulas that work well for a two-country organization break down at global scale. Exchange rate volatility, local market conditions, and local legal requirements (including the EU Directive's equal-value analysis) can all cause formula outputs to diverge from both market competitiveness and internal fairness. The formula becomes a source of systematic error rather than systematic fairness.

RSU cliffs create predictable attrition spikes. If you can reconstruct your organization's vesting calendar, you can predict when attrition risk is elevated: at the one-year mark (cliff decision), and at the four-year mark (full vest). These are not random — they are structurally induced by the standard vesting schedule. Organizations that do not proactively manage compensation competitiveness and engagement at these inflection points will experience predictable voluntary attrition.

Location-based pay interacts with the EU "equal value" framework. The Directive's equal value criteria — skills, effort, responsibility, working conditions — do not include geography. An employer paying a Senior Software Engineer in Warsaw 40% less than the same role in Amsterdam is making a business judgment the Directive may require them to justify. Location is not automatically a gender-neutral criterion.

The 5% trigger is lower than most orgs expect. Most organizations that have not run regression pay equity analysis assume they are well clear of a 5% gap. The actual pattern in tech is that base salary gaps are often within 5%, but total compensation gaps (including discretionary bonus, equity refresh, and signing bonuses) are frequently larger and more gendered. The Directive's reporting requirements will surface total remuneration, not just base salary.