Why a Higher Wage Floor Can Shrink Employee Equity
This chapter isolates the mechanism that makes employee equity in startups different from a simple salary substitute: when a firm is required to pay more cash, the equity part of a worker's package may contract. The key question is not whether founders prefer shares to wages in the abstract, but how a binding cash requirement changes a compensation contract that must satisfy both the employee's outside option and the startup's liquidity constraint. The Finnish evidence turns that question into an observable test.
The result may seem counterintuitive. If equity helps a startup conserve cash, why would a higher wage requirement lead to less equity rather than more? The answer is that equity is not an extra resource that can be added whenever cash becomes scarce. It is one of the ways the firm meets the employee's required package. When a larger share of that package must be paid immediately, there may be less room for a contingent claim whose value is uncertain and whose grant dilutes existing owners.
The contract has to clear two constraints
Consider a founder hiring one employee. The employee has an outside option: another job, self-employment, or simply the value of not accepting this particular risk. To accept the startup job, the expected value of cash wages and equity together must meet that outside option. The startup cannot offer a package that falls below it and expect the employee to remain willing to work.
Now add the founder's financial constraint. The startup has only so much liquid cash available for payroll. Its choice is not between an unlimited wage and an unlimited equity grant. It must select a wage that it can finance and an equity share that it is willing to issue, while preserving enough of the firm's future value and control for existing owners. In a young company, these limits can bind at the same time.
A wage floor changes the problem by putting a lower bound on the cash component. If the legal or economic minimum is below the wage the firm would otherwise pay, nothing changes. But if the floor is binding, the firm must raise cash compensation whether or not that is the way it would have preferred to satisfy the employee's participation condition. With the total package held to the level needed to meet the employee's outside option, the equity component becomes the residual. A higher required wage leaves a smaller residual for equity.
This is the intuition behind the Finnish paper's first prediction. In its parsimonious model, when the wage floor binds, optimal equity falls as the required wage rises. The model is not saying that an employee's total expected value must fall one-for-one with the equity reduction. It is showing why a constraint on immediate cash can change the composition of compensation, even when the firm still sees a reason to use equity.
The distinction between composition and total value matters. Suppose a startup must make a worker whole enough to accept a risky job. A higher cash wage may make the position more attractive in the short run, while a smaller equity grant reduces participation in future upside. The employee may prefer one combination depending on their financial needs, risk tolerance, and belief in the company. From the founder's perspective, the adjustment also changes dilution. Paying more cash consumes liquidity now; granting more equity gives away part of a future claim and may redistribute control. Neither instrument is costless, and neither cost is measured by the same clock.
The prediction applies only when the wage requirement actually binds. A firm already paying above the relevant floor need not change its package when the rule is extended. Even among exposed firms, workers can face different adjustments because the schedule is organized by occupation and experience. The difference between a worker's prior wage and the applicable schedule is therefore economically important: it determines how much new cash the firm must find, not merely whether a legal label has changed. A worker whose wage already clears the schedule may be unaffected, while one who is substantially below it may require a large immediate increase. This is why a single industry-wide change can produce different compensation responses inside the same startup.
That heterogeneity also clarifies what “substitution” means here. It does not require a founder to take a fixed dollar amount away from equity every time a wage rises. The firm could expand the total package, borrow, raise external capital, reduce hiring, or redistribute grants. The model isolates the case in which the startup must meet the employee's participation condition while operating under a binding cash constraint. In that case, the higher cash requirement consumes room that could otherwise have been used for equity. The empirical question is whether this predicted margin moves when an outside rule changes the cash requirement. It is not whether every employee experiences the same wage increase or every founder follows the same contract strategy.
A further distinction concerns who bears the uncertainty. A higher wage shifts more value into a payment the employee can use now and leaves more of the firm's future risk with the owners. A larger equity grant does the opposite: it gives the employee more exposure to the firm's eventual outcome and transfers some upside away from the founder. That transfer can be valuable if ownership strengthens effort or makes the employee willing to accept startup risk, but it may be a poor response when the firm needs to preserve control or when the employee cannot afford to wait. A wage floor can thus alter not only the accounting mix but also the allocation of risk between the people doing the work and the people who already own the company.
Equity also has an incentive channel that is separate from this financing choice. In the model, the employee's effort affects output, and the employee receives a share of that output through equity. A larger equity stake makes effort more valuable to the employee at the margin. That can raise effort, increase the value of remaining with the firm, and improve the chance of internal advancement. The claim therefore has two linked roles: it helps the startup manage cash and can influence behavior after the contract is signed.
Those roles can point in different directions. A firm may want equity because it cannot afford a high cash wage, yet a wage rule may force it to use more cash and issue less equity. It may still want the remaining equity to align effort or support retention. Conversely, a cash increase can improve an employee's immediate security even as it removes a future claim. Calling the change “less compensation” without identifying the form of compensation would miss the trade-off; calling it “equivalent compensation” would miss the risk and ownership transfer.
The model deliberately leaves out other reasons firms grant equity. Screening, risk sharing, tax treatment, and broader governance considerations may matter in actual contracts. The paper also notes that firms can respond to a cash shock by reallocating equity among workers, changing hiring, or using another financing margin. The negative response is therefore a prediction for a constrained compensation choice, not a mechanical rule for every employer.
The sequence has an important stopping point. The wage rule directly changes required cash pay; it does not directly order a founder to remove shares or command an employee to work harder. Equity is the endogenous response between the regulatory change and later behavior. That distinction is what allows the Finnish setting to test whether a cash constraint reshapes ownership allocation.
How Finland turns the model into a testable shock
The empirical setting is a statutory feature of Finnish collective bargaining. Agreements negotiated between unions and employer associations can be extended so that their occupation–experience wage schedules become legally binding across an industry, including firms that were previously outside the agreement. A startup that was already covered has little reason to experience a discrete change at the extension date. A previously uncovered, non-unionized startup may suddenly have to raise pay for workers whose pre-extension wages sit below the relevant schedule.
The important comparison is therefore not simply between industries with different wage levels. The study defines extension exposure using the post-extension status of the industry and the firm's pre-extension non-unionized status. Union status is measured before the regulatory event, so an individual startup does not choose its treatment in response to the new equity decision. The design uses the fact that the same industry-level decision has a different immediate cash consequence for firms that were already covered and firms that were not.
Exposure also varies within the workforce. The negotiated wage floors are not a single number for everyone; they depend on occupation and experience. Because the administrative data do not report the negotiated schedule directly, the study estimates a proxy: the modal annual wage bin among unionized firms in the same occupation, experience, industry, and year cell. For an employee in a non-unionized firm, the distance between their pre-extension wage and that modal schedule measures how far below the likely floor they were. A worker already at or above the schedule should face little mechanical adjustment. A worker substantially below it should face a larger required increase.
That worker-level difference strengthens the mechanism test. A firm-wide change associated with collective bargaining could affect all employees in some way. But a response that is larger for workers with a predetermined wage shortfall lines up specifically with the mechanical wage-floor channel. The exposure measure is not built from the worker's post-extension outcome; it uses the earlier wage, before the regulatory change. In principle, this lets the study ask whether the people facing the largest cash shock are also the ones around whom the compensation mix changes most in the predicted direction.
The study includes removals of extensions as well. A removal restores wage-setting discretion and therefore represents a relaxation of the binding constraint. But the two directions should not be expected to mirror one another perfectly. Wage increases imposed by an extension are immediate and legally binding. Removing the floor does not require a firm to undo a wage increase, recreate an old ownership allocation, or renegotiate a multi-period contract. Equity also reflects financing history and prior grants. The paper consequently treats extensions as the primary source of tightening and uses removals as complementary variation, not as proof that every response must reverse symmetrically.
The first-stage results are consistent with the model. Extension exposure reduces logged individual employee shares by about 0.114 log points in the reported baseline first stage. It also raises wages, and the interaction between extension exposure and wage-schedule distance produces strong within-firm variation in the wage response. The timing fits the institutional story: event-time estimates show no anticipatory movement in equity, adjustment begins around the extension, and the decline persists afterward. That is what one would expect if firms re-optimized compensation after a binding cash requirement rather than if equity had been drifting in anticipation of a known worker outcome.
There is still a difference between a plausible shock and a complete causal conclusion. Collective bargaining agreements contain provisions beyond wages, including rules concerning hours, leave, and dismissal. An extension could also coincide with changes in labor demand or firm survival. The study addresses these concerns with fixed effects, worker-level exposure, event-time tests, alternative specifications, and checks of reduced-form magnitudes. But the central logic remains conditional: the regulatory event must affect later exit and promotion through the compensation changes being instrumented, rather than through an unrelated path that happens to coincide with the wage schedule.
The mechanism also has a practical implication for how to read a smaller equity grant. It may not mean that the startup has abandoned incentives or that the employee has become less important. It may mean that a legal cash obligation consumed part of the compensation budget that would otherwise have been allocated as ownership. The resulting package can contain more immediate security and less contingent upside, while the firm still relies on the remaining equity to shape effort or retention. A change in the mix is not a simple change in the total amount of pay.
The next task is to determine whether this mechanism is genuinely responsible for the observed workforce effects. Employees who receive equity are not randomly selected, and firms that can afford or choose to grant it may differ from firms that cannot. The wage-floor setting supplies the variation needed to address that problem, but its credibility depends on the sample, the timing tests, and the exclusion assumptions. Only after those pieces are examined can the equity response be connected cautiously to staying and advancing.