A warehouse worker takes a job in one state and receives no paid sick days and no required rest break during a ten hour shift. She moves to a neighboring state, does the same job for a similar employer, and now the law guarantees paid sick time, a meal period, and a higher minimum wage. Nothing about her changed. The employer did not become more generous. What changed was the jurisdiction, and the jurisdiction matters in American employment law more than in almost any other field of federal legislation. That single experience generates an enormous volume of practical searches, and it exposes the central confusion this article resolves: most people assume federal law is the main source of workplace rights and that subnational law merely fills in details. For wages, safety, discrimination, and leave, that description is roughly right. For breaks, paid sick time, paid family leave, and protection against dismissal without cause, it is wrong in a way that has real consequences, because on those protections there is no federal rule at all and everything depends on the law of the state where the work is performed.

This article names that structure floors, ceilings and gaps. A floor is a federal rule that sets a minimum and preserves more protective state law. A ceiling is a federal rule that preempts state law so that the federal standard is the only one that can operate. A gap is a subject where federal law is silent and the worker’s rights are whatever the state has enacted, which in a state with no additional law means nothing at all. The organizing principle a reader should carry away is this: most federal labor and employment statutes are floors, a small number are ceilings, and on the questions people search for most often there is only the gap. Everything that follows is a tour of which statutes fall in which category, by their own savings and preemption provisions.

The three arrangements can be stated with the formality the subject deserves, because precision here prevents every common error below. A floor exists when a federal statute sets a minimum standard and contains language preserving more protective jurisdictional law. The federal standard binds every covered employer or worker, the jurisdictional standard binds where it exceeds the federal one, and the operative rule is always the more protective of the two. The four floors in this article are the wage and hour statute, the workplace safety statute, the civil rights employment title, and the leave statute, and each names its own savings language. A ceiling exists when a federal statute or doctrine displaces jurisdictional law so that the federal standard is the only one that can operate. The jurisdiction cannot add protections, cannot subtract them, and cannot route around the federal rule through clever drafting, as the deemer clause of ERISA section 514 demonstrates. The two ceilings in this article are the benefits statute’s express preemption and labor relations’ implied preemption. A gap exists when no federal statute addresses the subject at all. The jurisdiction’s law is the entire body of applicable law, and in a jurisdiction with no additional law the worker has no right on that subject. The four gaps in this article are meal and rest breaks, paid sick leave, paid family leave, and just-cause dismissal protection.

The One Test for the article as a whole follows from those definitions. A reader who has mastered the framework can state the organizing principle, that most federal statutes set floors jurisdictions may exceed while a small number preempt jurisdictional law entirely; can identify which is which for each major protection; and can state the practical consequence, that on several of the protections workers care about most there is no federal rule at all and the answer depends entirely on the jurisdiction. The namable claim is floors, ceilings and gaps: American employment law is not a federal system with jurisdictional supplements but three different arrangements operating at once, and knowing which of the three applies to a given protection answers most practical questions faster than reading any single statute.

One further distinction sharpens the framework before the tour begins. The express invitation of section 14(b) is neither a floor nor a ceiling nor a gap. It is a fourth posture, a federal authorization for the jurisdiction to move in the less protective direction, and it belongs in the analysis because it explains the most famous division in the field. The framework therefore has three main arrangements and one special case, and the special case is what makes the labor relations map look different from every other map in the article.

Can a state set a higher minimum wage than the federal government?

Yes. The wage and hour statute contains an express savings clause, 29 U.S.C. 218, stating that no provision of the federal law excuses noncompliance with any state law establishing a higher minimum wage. The federal rate is a floor, never a ceiling.

The Floor Rule and Its Four Statutes

The most common arrangement in this field is the floor. Four of the statutes that define American workplace rights share the same design. Each sets a national minimum. Each contains language that preserves state law doing more. Together they establish the default expectation that, where Congress has acted, a state may add protections on top of the federal standard but may not go below it. Understanding this design is the One Test for the floor category: a reader should be able to name the four statutes, the savings language in each, and the practical consequence that a state with stronger rules enforces the stronger rule while a state with no additional rules is governed by the federal floor alone.

The wage and hour statute is the oldest and clearest example. The Fair Labor Standards Act, as amended by the Fair Minimum Wage Act of 2007, Public Law 110-28, set the federal minimum wage at seven dollars and twenty five cents per hour effective July 24, 2009, in the third of three steps that began at five dollars and eighty five cents on July 24, 2007 and continued at six dollars and fifty five cents on July 24, 2008. The floor mechanism lives in section 18 of the Act, codified at 29 U.S.C. 218, whose first subsection provides that no provision of the Act shall excuse noncompliance with any federal or state law or municipal ordinance establishing a minimum wage higher than the federal minimum or a maximum workweek lower than the federal maximum, and that no provision of the Act relating to child labor shall justify noncompliance with any law establishing a higher standard. The Congressional Research Service has summarized the effect plainly: where a state enacts minimum wage, overtime or child labor law more protective of employees than the federal provision, the state law applies. The federal rate therefore functions only where the state has nothing higher to say. A state with a higher rate enforces the higher rate. A state with no higher rate leaves the federal rate as the operative one. The statute does not permit a state to set a lower rate than the federal minimum, because the floor runs only in one direction.

Why does the minimum wage differ across jurisdictions?

The difference is the savings clause working as designed. Section 218 preserves any higher state or municipal rate, so each jurisdiction’s effective minimum is the higher of the federal floor and whatever the jurisdiction enacted. Uniformity was never the design.

The wage floor repays closer study because it is the provision most readers meet first and misread most often. Three features of section 218 deserve attention. The first is its one-directional character. The clause preserves laws establishing a higher minimum wage, a lower maximum workweek, and a higher child labor standard. It does not preserve a lower wage, a longer workweek, or a weaker child labor rule. A jurisdiction that tried to set a minimum below the federal rate would find no shelter in section 218, because the federal Act would still bind every covered employer in that jurisdiction. The floor is therefore asymmetric by design: movement is permitted upward and forbidden downward. This asymmetry is what makes the federal rate a true minimum rather than a suggestion.

The second feature is the clause’s reach beyond state legislatures. Section 218(a) names federal law, state law, and municipal ordinance in the same sentence, which means cities and counties may set higher local minimums under the same authority that lets states do so. The local wage ordinances that multiplied in large cities rest on this language. An employer in a city with a higher local ordinance owes the local rate, an employer elsewhere in the same state owes the state rate if one exists, and an employer in a jurisdiction with neither owes the federal rate. The three layers stack without conflict because the rule at every layer is the same: pay the highest applicable rate.

The third feature is what the clause does not cover. Section 218 reaches the minimum wage, the maximum workweek, and child labor standards. It does not create a general savings clause for every provision of the Fair Labor Standards Act. Questions the Act leaves unaddressed, such as whether breaks must be provided, fall outside the floor entirely and belong to the gap category discussed later in this article. The floor is wide but not universal, and its boundaries are the boundaries of the subjects section 218 names.

The federal rate itself illustrates the floor in motion. The Fair Minimum Wage Act of 2007, enacted as part of Public Law 110-28 and signed May 25, 2007, raised the federal minimum in three steps: five dollars and eighty five cents effective July 24, 2007, six dollars and fifty five cents effective July 24, 2008, and seven dollars and twenty five cents effective July 24, 2009. Each step reset the floor nationwide, and each step left every higher state and local rate undisturbed, because a higher rate was already above the new floor. The Department of Labor’s wage and hour guidance states the current federal figure for work performed on or after July 24, 2009 as seven dollars and twenty five cents per hour. A reader who encounters that figure should understand it as the bottom of a staircase, not the level of the landing: it governs only where no higher law speaks, and the savings clause guarantees that every higher law survives it.

One structural contrast sharpens the point. The Fair Labor Standards Act contains no broad preemption clause comparable to the one ERISA carries. Congress did not need to write a ceiling into the wage statute because the floor design already allocates authority: the federal government sets the minimum, jurisdictions may exceed it, and nothing in the Act displaces the exceeding law. The recurring error of assuming the minimum wage is the same everywhere comes from treating the federal rate as the rule rather than the floor. Once the reader sees section 218 as the operative provision, the variation across jurisdictions stops looking like noncompliance and starts looking like the statute working exactly as written.

How the Floor Became the Default Drafting Technique

The floor did not have to be the default. Congress could have written each of these statutes as a ceiling, displacing jurisdictional law entirely in the manner of ERISA section 514. That it chose floors for wages, safety, discrimination, and leave reflects a legislative judgment about the subjects themselves, and understanding the judgment explains why the pattern is so consistent.

The judgment has two parts. The first is that these four subjects involve conditions that vary legitimately across the country. The cost of living differs between jurisdictions, which makes a single national minimum wage either too low for expensive jurisdictions or too high for inexpensive ones; a floor lets each jurisdiction calibrate upward. Workplace hazards differ by industry mix, which makes uniform safety enforcement less sensible than a baseline with jurisdictional programs built to local conditions. Conceptions of fair treatment evolve at different speeds, which makes a federal minimum catalog of protected characteristics with jurisdictional additions more workable than a frozen national list. Family needs and fiscal capacity differ, which makes unpaid federal leave with jurisdictional paid programs more passable than a single funded mandate. In each case the floor lets Congress act nationally without pretending that one rule fits every jurisdiction.

The second part of the judgment is political. A floor is easier to enact than a ceiling because it threatens fewer existing arrangements. A ceiling displaces every jurisdictional law on the subject, which mobilizes opposition from every jurisdiction that has one. A floor displaces only the substandard, which mobilizes almost no one, because few legislators defend the right to pay less than the federal minimum or to discriminate more than the federal title forbids. The savings clause is therefore not only a federalism device but a coalition-building device: it buys the votes of legislators who want stronger protections in their jurisdictions by promising that the federal statute will not undo them. The ceilings, by contrast, were enacted where Congress valued uniformity more than coalition breadth, in benefits administration across multi-jurisdiction employers and in the national labor policy the Board administers.

The drafting consequence is visible in the chronology. The wage floor dates to the Fair Labor Standards Act as amended, the safety floor to 1970, the discrimination floor to 1964, and the leave floor to 1993. Across five decades and very different political moments, Congress reached for the same technique whenever the subject involved variable local conditions and whenever the coalition needed the promise that federal action would not preempt stronger jurisdictional law. The consistency is the evidence for the judgment: the floor is the default because it solves the same legislative problem every time, how to set a national minimum without erasing the jurisdictions that want more.

Why didn’t Congress make every statute a ceiling?

Because ceilings are harder to pass and worse suited to subjects where conditions vary. A ceiling displaces every jurisdictional law on the subject, mobilizing opposition, while a floor displaces only the substandard and lets jurisdictions calibrate upward. Congress chose the technique that fit the subject and the coalition.

The second floor belongs to the workplace safety statute. The Occupational Safety and Health Act of 1970 created federal standards enforced by the Occupational Safety and Health Administration, but section 18 of the Act, codified at 29 U.S.C. 667, builds a state role into the structure. Under subsection 18(a), nothing in the Act prevents a state agency or court from asserting jurisdiction over a workplace safety issue where no federal standard under section 6 is in effect. Under subsection 18(b), any state that wants to take responsibility for safety issues where federal standards do exist must submit a state plan to the Secretary of Labor. Subsection 18(c) sets the approval test: the plan must designate a responsible state agency, provide standards and enforcement at least as effective as the federal standards for the same issues without unduly burdening interstate commerce, provide inspection rights at least as effective as section 8 with no advance notice, demonstrate legal authority and qualified personnel, and provide adequate funding. The phrase at least as effective is the operative floor. A joint 2015 letter from the National Institute for Occupational Safety and Health and the Occupational Safety and Health Administration reported 28 approved state plans that year, a figure that reached that number with the addition of a Massachusetts public-employee plan approved in August 2015, so the precise count on May 1, 2015 is treated here with the dated sourcing rather than as a count tied to that exact date. For readers tracing how that mechanism works in practice, the full history of the 1970 statute is developed in our account of the occupational safety and health act of 1970, which explains the state plan structure alongside the federal enforcement apparatus.

The third floor is the civil rights employment title. Title VII of the Civil Rights Act of 1964 bars employment discrimination on specified grounds, but section 708, codified at 42 U.S.C. 2000e-7, provides that nothing in the title shall be deemed to exempt or relieve any person from any liability, duty, penalty or punishment under any present or future law of any state or political subdivision, other than a state law that purports to require or permit an act that would be an unlawful employment practice under the title. The design is again a floor with one reservation. A state may prohibit discrimination by smaller employers than the federal title reaches. It may prohibit discrimination on additional grounds beyond those Congress listed. It may provide longer filing deadlines or stronger remedies. What a state may not do is require or permit an act the federal title makes unlawful, which is the narrow sense in which the federal rule also functions as a restraint. The practical result is that the anti-discrimination protections a worker can invoke depend on the jurisdiction. In a state with no broader fair employment law, the federal grounds and thresholds control. In a state with a broader one, the broader one governs.

The discrimination floor operates through an institutional mechanism that the other floors lack: deferral. Charges of employment discrimination filed with the Equal Employment Opportunity Commission are deferred to the state or local fair employment agency where one exists, under worksharing arrangements that divide investigative labor between the federal commission and its jurisdictional counterparts. The deferral system is the administrative expression of the floor. Because section 708 preserves the jurisdiction’s own law, the jurisdiction’s agency applies the jurisdiction’s broader statute, with its lower employer-size thresholds, its additional protected characteristics, and its longer filing windows, while the federal commission retains the federal claim. A worker who files in a jurisdiction with a strong fair employment law effectively pursues two claims at once, and the broader one sets the practical outcome.

Three dimensions of upward movement deserve separate notice because each answers a recurring practical question. The first is employer size. Title VII reaches employers above a statutory threshold, which leaves small workplaces outside the federal title. Section 708 permits jurisdictions to prohibit discrimination by smaller employers, and many have extended coverage downward to workplaces the federal threshold excludes. A worker at a small business has federal protection only if the jurisdiction supplied it. The second dimension is protected characteristics. The federal title lists specified grounds, and section 708 permits jurisdictions to add more. The federal list is therefore a minimum catalog, and the full catalog in any jurisdiction is the federal list plus whatever the jurisdiction added. The third dimension is procedure and remedy. Jurisdictions may set longer filing deadlines and provide stronger remedies than the federal title, which changes not only what is unlawful but how practically enforceable the right is.

The reservation in section 708 marks the one boundary on upward movement. A jurisdiction may not require or permit an act that would be an unlawful employment practice under the federal title. The reservation runs against state action that would authorize discrimination, not against state action that prohibits more of it. In practice the boundary rarely binds, because jurisdictions use their section 708 authority to expand protection rather than to authorize what the federal title forbids. But the boundary explains why the discrimination provision is described as a floor with one reservation rather than a pure floor: the federal rule sets the minimum and also polices the one direction in which a jurisdiction might move that Congress would not tolerate, which is downward into permission of federally prohibited discrimination.

Can a jurisdiction protect workers the federal threshold leaves out?

Yes. Section 708 of Title VII preserves state and local fair employment laws, so a jurisdiction may prohibit discrimination by employers too small for the federal title to reach and on grounds Congress did not list. The federal thresholds are a floor, not a ceiling.

The fourth floor is the leave statute. The Family and Medical Leave Act of 1993 provides certain employees with up to twelve weeks of unpaid, job-protected leave per year for qualifying family and medical reasons. Section 404 of the Act, codified at 29 U.S.C. 2654, contains two complementary rules. Subsection (a) provides that nothing in the Act affects the obligation of an employer to comply with any collective bargaining agreement or employment benefit program or plan that provides greater family or medical leave rights than the Act establishes. Subsection (b) provides the mirror image: the rights established under the Act shall not be diminished by any such agreement or plan. The implementing regulation, 29 CFR 825.701(a), extends the same principle to state law explicitly, stating that nothing in the FMLA supersedes any provision of state or local law that provides greater family or medical leave rights than the federal Act. Several states used that opening to create paid family leave programs, which is why the national rule is unpaid leave while some jurisdictions pay workers during part of the leave. A worker in a jurisdiction with no additional leave law has the federal twelve weeks, unpaid, and nothing more. A worker in a jurisdiction with a paid program has whatever the state added. Our separate guide to the family and medical leave act of 1993 develops the federal floor in full, including eligibility thresholds and the state paid programs that built on it.

The leave floor repays attention to the two subsections of 29 U.S.C. 2654, because they protect the federal right from two different directions. Subsection (a) provides that nothing in the Act affects the obligation of an employer to comply with any collective bargaining agreement or benefit program that provides greater family or medical leave rights than the Act establishes. This is the upward guarantee: a union contract or employer plan that gives more leave than the federal twelve weeks survives the Act untouched, and the employer must honor the greater promise. Subsection (b) provides the mirror image: the rights established under the Act shall not be diminished by any such agreement or plan. This is the downward guarantee: a contract cannot bargain away the federal minimum. Together the two subsections create a ratchet. Private ordering may add to the federal right and may not subtract from it, and the worker always holds the better of the two.

The regulation extends the same ratchet to public law. Section 825.701(a) of title 29 of the Code of Federal Regulations states that nothing in the FMLA supersedes any provision of state or local law that provides greater family or medical leave rights than the federal Act. The state paid family leave programs that several jurisdictions built rest on this regulatory extension of the statutory savings. The programs vary in design, but their legal foundation is identical: the federal statute guarantees unpaid time, the jurisdiction adds pay during part of that time, and section 2654 as implemented by the regulation preserves the addition. The result is the sharpest illustration in this article of why the floor matters more than the substantive rule. A reader who learns only that the FMLA provides twelve weeks of leave has learned the federal number. A reader who learns that the number is unpaid and that jurisdictions may add pay has learned the actual distribution of rights, which ranges from unpaid leave alone to partially paid leave depending on the jurisdiction.

The downward guarantee deserves emphasis because it answers a question workers ask in unionized workplaces: can the contract give less than the statute? Subsection (b) answers no. The federal twelve weeks cannot be bargained down, waived in a severance negotiation, or replaced by a less generous employer policy. The floor binds private agreements as well as jurisdictions, which makes the leave statute the most complete of the four floors: it constrains employers, unions, and legislatures alike from moving below the federal line, while inviting all three to move above it.

Does the federal leave law stop an employer from offering more?

No, and it stops no one from requiring more either. Section 2654 preserves greater rights in contracts, benefit plans, and state law alike. The federal twelve weeks of unpaid leave is a minimum that private parties and jurisdictions may exceed but never reduce.

Does the workplace safety law let states run their own programs?

Yes, through a state plan approved under 29 U.S.C. 667. The state must demonstrate standards and enforcement at least as effective as the federal standards, with adequate authority, personnel and funding. Where no federal standard exists, a state may assert jurisdiction directly.

The state plan mechanism deserves a closer look because it is the most procedural of the four floors, and its procedure is what makes the floor real. Section 18 divides the world into two situations. Subsection 18(a) addresses issues where no federal standard under section 6 is in effect: there, nothing in the Act prevents a state agency or court from asserting jurisdiction. This is the simpler half of the provision, a straightforward preservation of state authority over subjects the federal government has not occupied. Subsection 18(b) addresses the harder case, issues where a federal standard does exist. There a state that wants to take responsibility must submit a plan, and subsection 18(c) sets the approval test the Secretary applies.

The approval test has five elements, and the first two do the conceptual work. The state must designate a responsible agency, and it must provide standards and enforcement at least as effective as the federal standards for the same issues. The phrase at least as effective is the entire floor in four words. A state plan may innovate, may cover hazards the federal standards address thinly, and may enforce more aggressively, but it may not fall below the federal baseline on any issue the federal standards cover. The remaining elements secure the institutional capacity behind that promise: inspection rights at least as effective as section 8 with no advance notice, legal authority and qualified personnel, and adequate funding. A plan that promises strong standards without the inspectors to enforce them fails the test.

The mechanism also explains a structural fact about coverage. Federal OSHA does not cover state and local government workers, so the state plans, which must cover public employees, extend protection to a workforce the federal program leaves out. This is the floor operating at its most literal: the state program must be at least as effective as the federal one for covered private-sector issues and additionally reaches public workplaces the federal statute never entered.

Approval is not the end of federal involvement. The Secretary monitors approved plans against the at least as effective standard, and the statute provides for withdrawal of approval where a plan falls short. The floor is therefore a continuing obligation rather than a one-time certification. The dated count reported in this article, 28 approved plans in a 2015 joint agency letter, with Massachusetts added in August 2015, is stated with its date because the roster changes; the mechanism, a federal baseline with state programs permitted only at or above it, has been constant since section 18 was enacted.

Reading the Savings Clause: How Courts Tell a Floor From a Conflict

A savings clause does not end the legal analysis; it structures it. When a worker or an employer faces two rules, one national and one subnational, that point in different directions, the court works through a sequence that the savings clauses in this comparison make visible. Understanding the sequence is what turns the floor, ceiling or gap table from a summary into a tool.

The first question is whether the subnational rule is more protective than the federal one. Every savings clause in the floor set runs in one direction only. Section 218 of the wage statute preserves subnational laws establishing a higher minimum wage or a lower maximum workweek. Section 708 of the civil rights title preserves subnational laws that add liability, duty, penalty, or punishment. Section 404 of the leave statute preserves agreements and programs that provide greater leave rights. Section 18 of the safety statute approves state plans whose standards are at least as effective as the federal ones. In each case the comparison is substantive, not formal: the court asks whether the worker ends up with more protection, not whether the state used the same regulatory technique. A state that achieves a higher effective minimum wage through a different mechanism than the federal statute still satisfies the test, because the test measures the outcome for the worker.

The second question is whether the subnational rule contradicts the federal one in a way the savings clause cannot save. The Title VII provision states the limit expressly: a subnational law that purports to require or permit an act that would be an unlawful employment practice under the federal title is not saved. A hypothetical subnational law commanding employers to hire only members of a particular race would add liability in one sense and subtract civil rights in another, and the clause resolves the tension by defeating it. The same logic applies across the floor set. A state wage law that forbade overtime premiums the federal statute requires would not be more protective; it would be less, and the floor would reject it. A state safety plan that weakened federal protections could not be approved, because the “at least as effective” standard measures the plan against the federal level and finds it wanting. The floor is a one way ratchet, and the savings clause is the pawl that prevents backward motion.

The third question concerns the parties’ own agreements. The leave statute’s savings provision expressly reaches collective bargaining agreements and employment benefit programs that provide greater leave rights, which means a union contract can sit above the federal floor just as a subnational statute can. The implementing regulation then extends the principle to state and local law. The layered result is that the leave floor has three tiers: the federal minimum, the greater rights a contract or plan may supply, and the greater rights a state or locality may supply, with each higher tier controlling for the workers it covers. The wage statute works the same way in practice, though its text speaks of laws rather than contracts: a collective bargaining agreement that sets wages above the federal minimum is enforceable on its own terms, and the federal floor simply guarantees that no worker falls below it.

A final interpretive point deserves emphasis because it explains the drafting pattern across the four statutes. Congress wrote each savings clause in the negative: nothing in the statute shall excuse noncompliance, nothing shall be deemed to exempt, nothing shall diminish. The negative form matters. It means the federal statute never has to list the subnational laws it permits; it simply declines to displace them. The burden therefore falls on the party claiming preemption to show that the subnational rule falls outside the savings clause, not on the state to show that its rule fits inside. That allocation of the burden is part of what makes floors durable: the default is coexistence, and displacement is the exception that must be proved.

How does a court decide whether a subnational rule is more protective?

It compares the worker’s position under each rule and asks which leaves the worker better off. The comparison is substantive rather than formal: a different regulatory technique that produces a higher effective wage, a broader antidiscrimination duty, or a longer leave entitlement satisfies the savings clause.

The burden allocation has a practical consequence for litigation strategy. Because the default is coexistence, the employer who wants to defeat a subnational protection must do more than point to the federal statute; it must show that the subnational rule falls into one of the categories the savings clause does not protect. In wage litigation, that means showing the subnational rule is not in fact more protective, a difficult showing when the rule sets a higher number. In antidiscrimination litigation, it means showing the subnational rule requires or permits an unlawful employment practice, which succeeds only against the rare subnational law that commands discrimination rather than forbidding it. In benefits litigation, the burden runs the other way, because the ceiling is the default: the subnational government defending its mandate must show that the law does not relate to a plan, or that it is saved as insurance regulation without being defeated by the deemer clause. The allocation of the burden is therefore part of the arrangement itself. Floors place the burden on the challenger; ceilings place it on the defender. A reader who knows who must prove what knows how the dispute will likely end before it begins.

Who has to prove that a subnational rule survives?

On a floor, the challenger must show the subnational rule falls outside the savings clause, because coexistence is the default and displacement is the exception that must be proved. On a ceiling, the defender must show the rule escapes preemption, because exclusivity is the default and the exception must be proved instead.

The Ceiling: Where Federal Law Displaces the State

When does federal law block a state from adding protections?

Federal law blocks state action in two distinct ways. A statute may preempt expressly, by stating that it supersedes subnational law, as the benefits statute does. Or preemption may arise by implication from the structure of the federal scheme, as the labor relations doctrines hold.

The benefits statute carries the broadest express preemption clause in the field. Section 514 of the Employee Retirement Income Security Act of 1974, codified at 29 U.S.C. 1144, provides that the statute supersedes any and all subnational laws insofar as they may now or hereafter relate to any employee benefit plan covered by the Act and not exempt from it. The phrase “relate to” is the engine of the provision, and courts have read it expansively: a subnational law relates to a plan not only when it regulates plans directly but when it has a connection with or reference to them, which reaches far beyond statutes that mention benefits at all. The provision is a ceiling in the full sense. A state that wants to require employers to offer particular benefits, or to mandate particular terms inside benefit plans, collides with section 514 and loses. The full structure of the provision contains two refinements that sophisticated readers must know. The savings clause, section 514(b)(2)(A), preserves subnational laws that regulate insurance, banking, or securities, so a state insurance mandate directed at insurers rather than at plans can survive. The deemer clause, section 514(b)(2)(B), then prevents plans and their trusts from being deemed insurers, banks, or investment companies for purposes of state regulation, which stops states from using the insurance savings clause as a back door to regulate self funded plans by relabeling them. The three clauses together, preemption, savings, and deemer, form a coordinated architecture whose practical effect is that benefit plan content is a national subject. A worker whose dispute concerns the terms of an employee benefit plan will find that federal law, not the law of her state, supplies the controlling rules, and her subnational legislature cannot rewrite those rules no matter how protective it wishes to be. The complete guide to that statute walks the three clauses through the litigation that defined them.

Labor relations preemption works differently and reaches just as far. The National Labor Relations Act of 1935 contains no general express preemption clause for the conduct it governs. Instead the Supreme Court derived two doctrines from the structure of the Act, and together they occupy the field. The first, from San Diego Building Trades Council v. Garmon, 359 U.S. 236 (1959), holds that the Act impliedly preempts state regulation of conduct that is arguably protected by section 7 or arguably prohibited by section 8, deferring jurisdiction over such conduct to the National Labor Relations Board. The word “arguably” does real work: the state need not be certain the conduct falls under the federal statute; if the conduct is even arguably covered, the state must stand aside and let the Board decide. The second, from Lodge 76, International Association of Machinists and Aerospace Workers v. Wisconsin Employment Relations Commission, 427 U.S. 132 (1976), holds that the jurisdictions, and the Board itself, may not regulate conduct that Congress intended to leave unregulated and controlled by the free play of economic forces. The case concerned a union’s peaceful concerted refusal to work overtime during bargaining, conduct that was neither protected nor prohibited by the Act but that Congress had deliberately left to the parties’ economic weapons. The two doctrines divide the field between them. Garmon covers conduct the Act addresses; Machinists covers conduct the Act deliberately leaves alone. Between them they leave the jurisdictions almost no room to regulate labor relations conduct on their own authority. A state that wanted to create its own unfair labor practice rules, or to supplement the Board’s remedies with state penalties, would run into one or both doctrines. The ceiling in labor relations is therefore judge made rather than text made, but it binds the jurisdictions just as firmly as the express language of section 514.

The floor is the default, but the ceiling is the arrangement that produces the sharpest surprises, because a ceiling means the state legislature cannot act even where its voters want it to. The broadest preemption in the field belongs to the pension and benefits statute, and its three-clause structure deserves close study because it shows how Congress can close every door it means to close.

The Employee Retirement Income Security Act of 1974, section 514, codified at 29 U.S.C. 1144, begins with a preemption clause of remarkable breadth. Section 514(a) provides that the Act shall supersede any and all state laws insofar as they may now or hereafter relate to any employee benefit plan covered by the statute. The phrase relate to is the hinge of the entire provision, and courts have spent decades interpreting how far its reach extends. The structure then narrows through two exceptions. The savings clause, section 514(b)(2)(A), provides that nothing in the statute shall be construed to exempt or relieve any person from any state law that regulates insurance, banking or securities. The deemer clause, section 514(b)(2)(B), then limits the savings clause: neither an employee benefit plan nor any trust established under such a plan shall be deemed to be an insurance company or other insurer, bank, trust company or investment company for purposes of any state law purporting to regulate insurance companies, banks, trust companies or investment companies. The three clauses together create the field’s widest ceiling. A state cannot regulate employee benefit plans through its general laws because those relate to the plans. It can regulate through genuine insurance regulation, but it cannot capture a self-insured plan by deeming it an insurer. The consequence for a worker is that benefits law is the most nationalized corner of employment law, the area where state lines matter least. Our complete guide to erisa 1974 traces the full preemption doctrine, the litigation that shaped the relate to standard, and the narrow corridors where state law survives.

The three-clause structure of section 514 rewards the close reading this article’s decision procedure demands, because each clause answers a different question and the three together leave almost no room for state action. The preemption clause asks whether the state law relates to a covered plan. Courts have read that standard broadly to capture laws having a connection with covered plans, which means a state mandate directed at what plans must cover, how they must administer claims, or what remedies disappointed participants may seek is displaced. The breadth is deliberate. Congress in 1974 confronted employers operating pension and health plans across dozens of jurisdictions and concluded that subjecting plan administration to fifty different sets of mandates would defeat the statute’s purpose of encouraging plan formation through uniform federal regulation. Whether one agrees with that policy judgment, the mechanism is unmistakable: the relate to standard is the widest preemption formula in the field.

The savings clause then asks a narrower question: does the state law regulate insurance, banking, or securities? A state insurance code provision that regulates insurers as insurers survives, which is why the states retain authority over the business of insurance even in a field dominated by federal preemption. The corridor is real but bounded. It protects regulation of the insurance industry, not regulation of employee benefit plans through insurance-flavored language.

The deemer clause closes the remaining path. It provides that a plan or its trust shall not be deemed an insurance company or insurer for purposes of state insurance law. Without this clause, a state could have captured self-insured employer plans by declaring them insurers and regulating them under the saved insurance authority, which would have reopened through definition what the preemption clause closed by substance. The deemer clause is therefore best understood as an anti-circumvention device: it prevents the savings clause from swallowing the preemption clause. An employer that self-insures its health plan operates under federal law; an employer that buys coverage from a commercial insurer operates under a plan whose insurer is subject to state insurance regulation. The worker’s protections differ between those two funding arrangements, and the difference is entirely the product of the three clauses.

The practical consequence is the one the table records. On wages, safety, discrimination, and leave, the worker’s rights are the better of the federal rule and the jurisdiction’s additions. On benefits, the worker’s rights are the federal rule, with the jurisdiction able to act only through genuine insurance regulation of insured arrangements. This is why benefits disputes are litigated in federal court under federal standards while wage disputes so often turn on which jurisdiction’s higher rate applies. The ceiling does not merely limit the jurisdiction; it reroutes the entire body of law into the federal forum.

Why does benefits law vary less across jurisdictions than wage law?

Because Congress wrote the two statutes with opposite federalism designs. The wage statute’s savings clause at 29 U.S.C. 218 preserves higher jurisdictional rates, producing variation. The benefits statute’s preemption clause at 29 U.S.C. 1144 displaces jurisdictional mandates relating to plans, producing uniformity. The arrangement, not the subject matter, explains the difference.

A second and subtler ceiling operates in labor relations. The National Labor Relations Act contains no broad statutory preemption clause comparable to ERISA’s, but the Supreme Court has built two implied preemption doctrines from the structure of the Act. The first, from San Diego Building Trades Council v. Garmon, 359 U.S. 236 (1959), holds that the Act impliedly preempts state regulation of conduct that is arguably protected by section 7 or arguably prohibited by section 8, deferring jurisdiction to the National Labor Relations Board’s primary authority over those questions. The second, from Lodge 76, International Association of Machinists v. Wisconsin Employment Relations Commission, 427 U.S. 132 (1976), holds that states may not regulate conduct that Congress intended to leave unregulated and controlled by the free play of economic forces, such as a union’s peaceful concerted refusal to work overtime during bargaining. Together Garmon and Machinists create a ceiling over the core of collective bargaining: where Congress protected concerted activity, states cannot restrict it; where Congress deliberately left the parties to economic combat, states cannot referee it either. The doctrines are implied rather than written, which makes them harder to read than an express clause, but their practical force is the same. They explain why a single national labor statute produces a uniform framework for the questions Congress assigned to federal control.

The two doctrines need to be distinguished carefully because they preempt for opposite reasons, and confusing them produces exactly the misreading this article is designed to prevent. Garmon preemption is about forum. It holds that where conduct is arguably protected by section 7 or arguably prohibited by section 8, the question belongs to the National Labor Relations Board in the first instance, and state courts and agencies must stand aside. The rationale is institutional: Congress created an expert federal tribunal to develop a uniform national labor policy, and permitting state tribunals to decide the same questions would fragment that policy across jurisdictions. The doctrine’s trigger word is arguably. The state need not be certain the conduct falls within the Act; if the conduct is arguably within it, the Board decides. A state court injunction against peaceful picketing that is arguably protected concerted activity is the classic casualty of Garmon: the state court may be right or wrong about the merits, but it is the wrong forum to decide.

Machinists preemption is about substance rather than forum. It holds that where Congress deliberately left conduct unregulated, to be resolved by the free play of economic forces, neither the states nor the Board may regulate it. The rationale is substantive: Congress struck a balance between labor and management by leaving certain economic weapons available to both sides, and state regulation would upset that balance even if the state purported to be neutral. The Machinists case itself involved a union’s peaceful concerted refusal to work overtime during bargaining, conduct the Act neither protected nor prohibited. The state labor board could not enjoin it, not because the Board had primary jurisdiction, but because no government was meant to decide the question at all. The parties were to fight it out economically.

The two doctrines together create a ceiling with a distinctive shape. Garmon covers the territory Congress regulated, assigning it to the federal tribunal. Machinists covers the territory Congress deliberately left alone, assigning it to no tribunal. Between them they leave the states almost no room in the core of collective bargaining: the state cannot decide what the Board should decide, and cannot decide what Congress decided no one should decide. What remains for the states in labor relations is the periphery the Act never touched, matters of general applicability like the contract and property law that governs every workplace, plus the express authorization of section 14(b) discussed next. The ceiling is implied rather than written, which is why readers miss it, but its practical force equals any express clause. A state legislature that wants to regulate the conduct of a bargaining campaign will find Garmon or Machinists waiting, and the doctrines do not yield to good intentions.

Can a state court enjoin a peaceful strike?

Not where the conduct is arguably protected or prohibited by the federal labor statute. Garmon defers such questions to the National Labor Relations Board’s primary jurisdiction. And where Congress left the conduct to economic forces, Machinists bars state regulation too. The state forum is displaced in both cases, for opposite reasons.

The floor and the ceiling exhaust the arrangements where federal law speaks at all. But the arrangement that generates the greatest volume of practical confusion is the third one, and it is the subject of the next section.

Why the Two Ceilings Are Built Differently

The two ceilings differ in their construction, and the difference matters for anyone trying to move them. ERISA preemption is express: Congress wrote the supersession into the statute, and the text names its own scope with the phrase “relate to.” Labor relations preemption is implied: Congress wrote no general preemption clause into the National Labor Relations Act, and the Supreme Court derived the doctrines from the structure of the scheme. The practical consequence is that the ERISA ceiling can be read on the face of the statute, while the labor relations ceiling must be assembled from Garmon, Machinists, and the cases that applied them.

The difference also changes what a subnational legislature faces. A legislature considering a benefit mandate can read 29 U.S.C. 1144 and see the barrier in plain text; the legal question is whether its proposal relates to a plan, a question the “relate to” standard answers broadly. A legislature considering a labor relations measure faces a harder inquiry: it must ask whether the conduct it wants to regulate is arguably protected or prohibited by the federal Act, or instead belongs to the zone Congress left to economic forces, and either answer leads to preemption through a different doctrine. The first barrier is textual and the second is structural, but both are ceilings in the full sense, because neither can be moved by the state acting alone.

The labor relations ceiling has one more structural feature the benefits ceiling lacks: an agency with primary jurisdiction. The National Labor Relations Board does not merely enforce the federal statute; under Garmon it is the exclusive forum for disputes over conduct arguably covered by the Act, which means the ceiling operates procedurally as well as substantively. A subnational court that tried to decide such a dispute would not only apply the wrong law but sit in the wrong forum. The benefits ceiling has no equivalent institution. ERISA disputes are heard in federal and subnational courts alike, applying federal law, because the preemption is substantive rather than jurisdictional. The difference matters for the reader’s practical question of where to file: a labor relations claim belongs before the Board, while a benefits claim belongs in court under federal law. Two ceilings, two procedural consequences, one lesson that the architecture of preemption determines the forum as well as the rule.

Why did Congress choose express preemption in one field and leave the other to the courts? The benefits statute was written in 1974 against a background of state insurance regulation that Congress knew intimately, and the drafters wanted no doubt about which sovereign governed plan content. The labor relations statute was written in 1935, when the modern preemption vocabulary barely existed, and Congress created a specialized agency, the National Labor Relations Board, to develop the law case by case. The Court’s doctrines grew around that agency’s primary jurisdiction. The two histories explain the two architectures: one ceiling was built by the legislature with a hammer, the other by the judiciary around an agency with a chisel. A reader who understands this difference understands why the ERISA ceiling looks the same in every decade while the labor relations ceiling has been refined, narrowed, and reaffirmed across generations of decisions.

What makes a ceiling harder to move than a floor?

A floor invites the subnational legislature to act, so change requires only a subnational statute. A ceiling forbids the subnational legislature from acting, so change requires Congress to amend the federal provision or the courts to narrow the doctrine, both of which are rarer and slower than state legislation.

Preemption in Practice: What the Doctrines Do to Real Disputes

The ceiling doctrines are abstract until they are applied, so this section works each one through the kind of dispute it actually resolves. The examples are illustrative, built from the doctrines’ settled elements rather than from any single case, and they show why the ceilings bind as firmly as any statutory text.

Take the benefits ceiling first. A subnational legislature, concerned about workers who lose health coverage when they change jobs, enacts a statute requiring every employer with more than fifty workers to offer a health plan meeting specified minimum standards. The statute does not mention any particular plan; it regulates employers. Under ERISA section 514, the law is nevertheless superseded, because it relates to employee benefit plans: an employer can only comply by establishing or modifying a plan, and the Supreme Court’s “relate to” jurisprudence reaches laws that dictate plan content even when they are framed as employer mandates. The subnational legislature’s next move is to direct the mandate at insurers instead, requiring every health insurer in the jurisdiction to offer a specified minimum policy. The insurance savings clause preserves this version, because it regulates insurance, but only as applied to insured plans. An employer with a self funded plan, which pays claims from its own assets rather than buying insurance, invokes the deemer clause: the plan cannot be deemed an insurer, so the insurance mandate cannot reach it. The subnational legislature has now exhausted its options. The insured market can be regulated through the savings clause, the self funded market cannot be reached through the deemer clause, and the original employer mandate falls to the preemption clause. Three clauses, one result: plan content is a national subject.

Take the labor relations ceiling next. A subnational legislature, concerned about aggressive picketing, enacts a statute creating a subnational unfair labor practice for recognitional picketing and authorizing subnational courts to enjoin it. A union pickets an employer for recognition, the employer sues under the subnational statute, and the union argues that the conduct is arguably protected or prohibited by the federal Act. Under Garmon, the subnational court must yield: the conduct is at least arguably covered by sections 7 and 8, the Board has primary jurisdiction, and the subnational injunction cannot issue. The subnational legislature tries again with a statute regulating the economic weapons of a lawful strike, limiting the duration of picket lines during contract negotiations. The conduct is neither protected nor prohibited by the federal Act; Congress left the parties’ economic weapons to the free play of economic forces. Under Machinists, the subnational regulation is preempted anyway, because the zone Congress left unregulated is as much a federal choice as the zone it regulated. The subnational legislature has again exhausted its options. Conduct the Act covers goes to the Board; conduct the Act deliberately leaves alone stays with the parties. The two doctrines divide the field between them and leave the subnational government outside both halves.

The practical lesson for the reader is that ceilings are not theoretical. They decide real disputes in favor of federal exclusivity, and they do so through mechanisms, the “relate to” standard, the “arguably” trigger, the unregulated zone, that are broader than a casual reader expects. The floor statutes invite the subnational legislature to the table. The ceiling doctrines remove the chair.

The contrast with the floor statutes is the point. When Congress wanted the jurisdictions to add protections, it wrote savings clauses in plain terms, as sections 218, 667, 2000e-7, and 2654 show. When it wanted the jurisdictions excluded, it wrote or left room for preemption, as sections 1144 and the labor relations doctrines show. The choice was never an accident of drafting. It was a substantive decision about where uniformity was worth more than experimentation, and the next section takes up the one case where Congress split the difference inside a single statute.

The Express Invitation: When Federal Law Authorizes Less Protection

One provision in the series stands alone as the clearest instance of a federal statute authorizing a state to move in the less protective direction, and it explains the famous split in American labor regimes. Section 14(b) of the National Labor Relations Act, codified at 29 U.S.C. 164(b), provides that nothing in the Act shall be construed as authorizing the execution or application of agreements requiring membership in a labor organization as a condition of employment in any state or territory in which such execution or application is prohibited by state or territorial law. The text addresses agreements requiring membership as a condition of employment, the union shop; in practice it is the statutory basis for the state laws known as right-to-work laws, which prohibit compelling workers to join a union or pay fees as a condition of keeping their jobs, including the agency-fee arrangements that some states also bar under this authority.

The significance of section 14(b) for the floor, ceiling and gap framework is that it inverts the usual relationship. The floors described above invite states upward. The ceilings forbid state action entirely. Section 14(b) invites states downward, expressly permitting them to prohibit what the federal statute would otherwise allow. That is why a single national labor statute produces two very different regimes. In jurisdictions without such laws, unions may negotiate security agreements that require financial support from the workers they represent. In jurisdictions with them, those agreements are void, and the union must represent all workers in the bargaining unit while collecting support only from volunteers. Whether one regards that outcome as protecting individual freedom or weakening collective power, the mechanism is statutory and explicit: Congress built the state option into the federal law, and the divergence across jurisdictions is the intended product of that design rather than an accident of it. The sequence in which these labor statutes accumulated, and the political settlements behind each one, is traced in our account of us labor legislation since 1950.

The 1947 compromise behind section 14(b) explains why the provision takes the unusual form of an express authorization to do less. The Taft-Hartley amendments of 1947 rewrote the 1935 labor statute to address what its drafters saw as union overreach, and the union security question was among the most contested. A union shop agreement requires workers to join the union as a condition of employment; an agency shop requires financial support without formal membership; an open shop requires neither. The federal statute as amended permitted the first two where state law did not forbid them, and section 14(b) is the clause that preserved the state’s power to forbid them. The compromise satisfied two constituencies at once: supporters of a national labor policy got a uniform federal framework for bargaining, and supporters of state control over the union security question got an explicit carve-out. Neither side got everything, which is the usual signature of a durable settlement.

The textual narrowness of section 14(b) deserves the precision the verification record requires. The provision speaks of agreements requiring membership in a labor organization as a condition of employment. Read on its own terms, it covers the union shop. In practice it has served as the statutory basis for state laws that also bar compulsory fee arrangements, the agency shop, on the understanding that compelled financial support raises the same policy question the provision addresses. A reader who quotes the text should quote it accurately and a reader who describes the practice should describe the gloss accurately; the two are related but not identical, and the distinction matters when the question is what exactly Congress authorized.

The result is the only downward invitation in the series. Every floor in this article invites the jurisdiction upward. Every ceiling forbids jurisdictional action. Section 14(b) alone invites the jurisdiction to prohibit what the federal statute would otherwise permit, which is why the national labor statute yields the most visibly divided map in American employment law. The division is not a failure of uniformity but the product of a Congress that chose uniformity for bargaining conduct and jurisdictional choice for union security, and wrote both choices into the same Act.

Is at will employment the federal rule for firing?

The federal government never created a general just-cause requirement, and the default employment relationship remains at will, meaning most workers may be discharged for any reason that is not independently unlawful. The only federal constraints are statutes barring specific illegal reasons for discharge, such as discrimination or retaliation.

The Historical Arc: How Three Arrangements Accumulated

The three arrangements did not emerge from a single design. They accumulated across eight decades, each statute responding to the political conditions of its moment, and the patchwork the table records is the sediment of those separate decisions. Walking the sequence shows why no single theory of the federal role ever took hold.

The story opens in 1935 with the National Labor Relations Act, the Wagner Act, which created the federal labor relations regime and the Board that administers it. The Act contained no general preemption clause, because the modern vocabulary of preemption had not yet developed, and the ceilings in the field grew later through the Court’s doctrines. Twelve years later, the Taft-Hartley Act of 1947 amended the Wagner framework and added section 14(b), the express invitation for subnational right to work laws. The invitation was a political compromise: the federal statute would permit union security agreements as a national baseline, but the opponents of compulsory unionism extracted the right of each subnational legislature to forbid them. The two regimes that resulted have persisted ever since, a living demonstration that a single federal statute can sustain opposite subnational outcomes when Congress writes the permission explicitly.

The wage floor arrived in 1938 with the Fair Labor Standards Act, the last major statute of the New Deal’s first wave. Its savings clause reflected a constitutional caution of the era: the Supreme Court had upheld wage regulation in its 1937 decisions, ending the Lochner era’s restrictions on such laws, and the drafters preserved subnational authority partly to avoid constitutional doubt and partly from a substantive belief that wage levels should reflect local conditions across a continental economy. The caution hardened into a permanent feature. Every subsequent federal minimum wage increase, culminating in the Fair Minimum Wage Act of 2007 that set the 7.25 dollar rate effective July 24, 2009, left the savings clause intact, so the floor ratcheted upward at the federal level while the subnational supplements continued to operate above it.

The antidiscrimination floor came in 1964 with Title VII of the Civil Rights Act. Its savings clause reflected a different politics: the civil rights coalition needed the votes of legislators who feared federal displacement of existing subnational fair employment laws, some of which predated the federal title and some of which were stronger. Section 708 bought those votes by promising that the federal statute would add to, not subtract from, subnational protections. The promise held, and the subnational laws that the clause preserved became the testing ground for protections the federal title did not yet reach.

The safety floor and the benefits ceiling arrived four years apart and illustrate the two directions Congress could move. The Occupational Safety and Health Act of 1970 created the state plan mechanism, inviting subnational enforcement under federal supervision. The Employee Retirement Income Security Act of 1974 created the broadest express preemption clause in the field, removing benefit plan content from subnational authority almost entirely. The contrast between the two statutes, enacted within a single decade, shows that the choice between floor and ceiling was never ideological in any simple sense. It was functional: safety enforcement could be delegated under supervision, while benefit plan administration across subnational lines could not tolerate fifty different content mandates.

The leave floor closed the sequence in 1993 with the Family and Medical Leave Act, the first statute in the set enacted after the modern conservative era began. Its savings clause was uncontroversial by then; the floor arrangement had become the default for new protective legislation, and no one proposed preempting subnational leave law. The gaps, meanwhile, are defined by what did not happen: no federal break statute, no federal paid sick leave law on the reference date, no federal paid family leave, no federal just cause statute. The historical arc therefore ends where the comparison began, with a statute book that invites the subnational governments to do more, forbids them from doing anything in two fields, authorizes them to do less in one provision, and says nothing at all on the protections workers ask about most.

The Four Gaps: Where Federal Law Says Nothing

The gap is the arrangement that surprises most readers, and it deserves the careful statement that negative facts require. For a large share of the protections workers actually invoke, including breaks, paid sick time, paid leave and dismissal protection, there is no general federal requirement. The accurate description is that state law does the majority of the work in the areas people ask about most, which is also why employment rights differ so sharply across jurisdictional lines.

The first gap is meal and rest breaks. The Wage and Hour Division has stated plainly that federal law does not require lunch or coffee breaks. Federal regulation addresses only the pay consequences of breaks that an employer chooses to offer: short rest periods of five to twenty minutes must be paid if offered, while bona fide meal periods, typically thirty minutes or more during which the employee is fully relieved of duties, are not compensable. Those are rules about pay for breaks, not rules requiring breaks to be provided. That gap must be stated with one verified caveat. Section 7(r) of the Fair Labor Standards Act, codified at 29 U.S.C. 207(r) and added by section 4207 of the Affordable Care Act effective March 23, 2010, requires employers to provide reasonable break time for a non-exempt employee to express breast milk for her nursing child for one year after the child’s birth, each time she needs to, plus a private place other than a bathroom, shielded from view and free from intrusion. Those breaks are unpaid, employers with fewer than fifty employees are exempt if compliance would impose an undue hardship, and the provision does not preempt more protective state law. The nursing-mother provision is therefore a narrow, targeted exception that proves the general rule: outside it, no federal statute requires meal or rest breaks, and whether a worker gets one is determined entirely by the law of the jurisdiction.

The second gap is paid sick leave. On May 1, 2015, no federal statute required employers to provide paid sick leave. In February 2015 Representative Rosa DeLauro introduced the Healthy Families Act, H.R. 932 with a Senate companion S. 497, which would have created a national paid-sick-days standard; the measure saw no action in the House subcommittee. The date-wall treatment applies to the one related development that postdates the reference frame of this article: Executive Order 13706, signed September 7, 2015, established paid sick leave for federal contractors only, and because it came after May 1, 2015 and covers only contract work for the federal government, it is not part of the law this article describes and is not a general federal paid-sick-leave requirement. The first general federal paid-sick-leave requirement of any kind, and it was temporary, came with the Families First Coronavirus Response Act effective April 1, 2020 and sunset December 31, 2020. At the date this article describes, the workers who had paid sick leave had it because their state or locality enacted it, and workers in jurisdictions without such laws had none.

The third gap is paid family leave. The federal leave statute provides up to twelve weeks of unpaid, job-protected leave per year, and the word unpaid is doing the load-bearing work in that sentence. The Department of Labor describes the FMLA as providing unpaid leave, and health policy analysis has confirmed that the 1993 Act mandates only unpaid leave. Several states built paid programs on the FMLA floor using the express savings of 29 U.S.C. 2654, which is why some jurisdictions pay workers during part of the leave while the federal baseline pays nothing. But in a jurisdiction with no additional law, the worker’s leave right is twelve weeks, unpaid, and job-protected, nothing more. The recurring error of assuming that the federal government guarantees paid family leave dissolves on contact with the statute: the guarantee is of time off, not of pay during it.

The fourth gap is just-cause protection against dismissal. The United States does not generally provide just-cause protections for employees; most workers are employed at will and may be terminated for any reason that is not otherwise unlawful. No federal statute imposes a general just-cause requirement. The federal constraints on discharge are prohibitions on specific illegal reasons for firing, such as discrimination barred by Title VII or retaliation barred by various whistleblower and rights provisions, not a requirement that the employer prove cause. This negative fact rests on the at-will default doctrine plus the absence of any federal just-cause statute, and it should be stated as such rather than attributed to a particular provision. One state, Montana, maintains a general just-cause requirement after an initial probationary period under its Wrongful Discharge from Employment Act; everywhere else, the protection against being fired without cause exists only if the jurisdiction created it.

What is right to work under the federal labor statute?

Section 14(b), 29 U.S.C. 164(b), expressly permits states to prohibit agreements requiring union membership as a condition of employment. In practice that authority underlies state right-to-work laws, which void union security and agency-fee arrangements so that no worker can be compelled to pay a union to keep a job.

The floor, ceiling or gap table

The findable artifact of this article is a single table. For each major protection it records whether the federal statute sets a floor, imposes a ceiling through preemption, or is silent, and what that means for a worker in a jurisdiction with no additional law.

Protection	Federal arrangement	Meaning for a worker in a jurisdiction with no additional law
Minimum wage	Floor: FLSA section 18, 29 U.S.C. 218	The federal rate of seven dollars and twenty five cents per hour, set by the 2007 Act effective July 24, 2009, governs
Overtime and child labor	Floor: FLSA section 18, 29 U.S.C. 218	The federal maximum workweek and federal child labor standards govern
Workplace safety	Floor: OSH Act section 18, 29 U.S.C. 667	Federal standards enforced by federal OSHA govern; state may run an approved plan at least as effective
Employment discrimination	Floor with one reservation: Title VII section 708, 42 U.S.C. 2000e-7	The federal grounds, thresholds and remedies govern; the jurisdiction may add grounds but may not permit what the title forbids
Family and medical leave	Floor: FMLA section 404, 29 U.S.C. 2654	Up to twelve weeks of unpaid, job-protected leave govern; no pay guarantee exists
Employee benefit plans	Ceiling: ERISA section 514, 29 U.S.C. 1144	Federal law supersedes state laws relating to covered plans, with narrow insurance, banking and securities exceptions
Labor relations, protected and prohibited conduct	Ceiling: implied preemption under Garmon, 359 U.S. 236 (1959), and Machinists, 427 U.S. 132 (1976)	Federal Board jurisdiction governs; the jurisdiction cannot regulate conduct Congress assigned to federal control
Union security agreements	Express state authorization: NLRA section 14(b), 29 U.S.C. 164(b)	The jurisdiction may prohibit membership-as-a-condition-of-employment agreements, producing a distinct regime
Meal and rest breaks	Gap, with a narrow federal exception for nursing mothers, FLSA section 7(r), 29 U.S.C. 207(r)	No general break requirement exists; the worker has no break right under federal law
Paid sick leave	Gap	No federal paid sick leave requirement exists; the worker has none under federal law
Paid family leave	Gap	The federal rule is unpaid leave; no federal pay guarantee exists
Just-cause protection against dismissal	Gap	The at-will default governs; no federal just-cause requirement exists

The table is deliberately conservative about numbers. State-by-state figures change, and this series states no current jurisdiction-by-jurisdiction data that shifts with each legislative session. The mechanism column is the durable part: the arrangement for each protection has not changed since the statutes were enacted, and the arrangement is what answers the practical question.

A Worker’s Week Under Three Arrangements

The comparison becomes concrete when it is walked through a single worker’s week, because the three arrangements do not operate in separate worlds. They operate on the same person, in the same job, in the same pay period, and the worker moves between them without noticing the jurisdictional seams. Consider a retail employee working forty five hours in a week for an employer with two hundred workers, in a state that has enacted a higher minimum wage, a paid sick days law, and broader antidiscrimination protections, but no meal period requirement and no right to work law. Every line of her experience illustrates one of the three arrangements.

Her wage is a floor question. The federal statute sets 7.25 dollars per hour and requires time and a half for the five hours beyond forty. The subnational statute sets a higher hourly rate. Section 218 of the federal statute preserves the higher state rate, so the employer owes the state rate for all forty five hours and the federal overtime premium structure for the hours beyond forty, unless the state also improves the overtime rule. The worker’s paycheck is therefore a composite: the federal statute supplies the overtime architecture and the subnational statute supplies the higher base rate, and neither sovereign’s rule displaces the other.

Her safety protections are a floor by substitution. If her state operates an approved plan under 29 U.S.C. 667, the inspectors who walk the stockroom are state employees applying state standards that the federal agency has certified as at least as effective as its own. If a hazard injures her, the citation the employer receives comes from the state program. The federal statute’s contribution is the effectiveness test the state had to pass years earlier, invisible in the daily operation but load bearing underneath it.

Her protection against discrimination is a pure floor. The federal title bars discrimination on its enumerated grounds for employers of her size, and the subnational law adds the additional grounds the subnational legislature chose. If she is harassed on a ground the federal title does not reach, her claim lives or dies under subnational law, and the federal savings clause at 42 U.S.C. 2000e-7 is what keeps that state claim alive against any argument that Congress occupied the field.

Her benefits are a ceiling. If she disputes the denial of a claim under the employer’s health plan, the state insurance rules she might expect to help her are superseded by ERISA section 514 wherever they relate to the plan, and the deemer clause prevents the plan from being treated as an insurer to bring those rules back in. Her subnational legislature’s views on benefit mandates are irrelevant to her dispute, because Congress removed the field from state authority in 1974.

Her organizing rights are a ceiling of the second type. If she and her coworkers discuss unionization, the Garmon doctrine means any state regulation of that conduct must yield to the National Labor Relations Board’s primary jurisdiction, and the Machinists doctrine means the state cannot regulate the economic weapons the parties use even where the federal statute is silent. Her state cannot create a parallel labor board with broader remedies, because the Supreme Court’s doctrines close the door the statute left unmarked.

Her union security obligation is the express invitation. Because her state has no right to work law, section 14(b) lies dormant, and federal law permits her employer and a union to agree that she must join or pay as a condition of employment. Had she taken the same job across the border in a right to work state, the same federal statute would produce the opposite result, with no change in the federal text.

Her breaks and her sick days are gaps with opposite outcomes. Her state has no meal period law, so the federal silence on breaks means she has no legal right to a lunch break at all, and the only federal break provision, the nursing mother rule, does not apply to her situation. But her state does have a paid sick days law, so the federal silence on paid sick leave means the subnational law is the entire law on the subject, and she accrues the paid sick time the state guarantees. Two gaps, two different answers, both determined by what the subnational legislature did with the empty federal space.

Her leave rights are a floor with a gap inside it. If she needs time off for a serious health condition, the federal statute guarantees up to twelve weeks of unpaid, job protected leave, and her jurisdiction’s law may add longer leave or paid benefits on top; the federal savings clause at 29 U.S.C. 2654 protects whatever the jurisdiction added. But the paid portion is a gap: the federal statute never created it, so if her jurisdiction had not enacted a paid program, her leave would be entirely unpaid no matter how clearly she needs the income. One protection, two arrangements stacked inside it, the unpaid floor from Washington and the paid supplement from the jurisdiction, or nothing.

Her protection against dismissal is the starkest gap. Her employer may end her employment for any reason not prohibited by a specific statute. The antidiscrimination floor and the antiretaliation provisions supply the specific prohibitions, but no general just cause rule exists in either sovereign’s law. If she is fired for a reason that is merely unfair, the three arrangements offer her nothing, because the protection she wants was never enacted by anyone.

One worker, one week, all three arrangements plus the express invitation, operating simultaneously. The map is complicated, but it is not random. Each line of her experience traces back to a specific congressional choice about the role of the states, and the floor, ceiling or gap table is the index to those choices.

The map also has a temporal dimension the table flattens. The arrangements describe the law as it stood on the May 1, 2015 reference date of this article, and each arrangement can change, but the mechanisms of change differ. A floor can be raised by a subnational statute at any time, and the savings clauses automatically accommodate the new higher standard without any federal action. A gap can be filled the same way, and the history of paid sick days and paid family leave is largely the history of subnational legislatures filling gaps one by one. A ceiling can only be moved by Congress amending the federal provision or by the courts narrowing the doctrine, which is why the benefits ceiling has looked the same for decades while the floors and gaps have shifted steadily. The express invitation of section 14(b) can be taken up or left alone by each subnational legislature independently, which is why the right to work map changes one jurisdiction at a time. The three arrangements are therefore not only different in space but different in time: floors and gaps evolve continuously, ceilings evolve rarely, and the invitation evolves jurisdiction by jurisdiction.

The Local Layer: Cities, Counties, and the Savings Clauses

The comparison has spoken of two sovereigns, but the savings clauses name three. The wage provision preserves municipal ordinances alongside subnational law. The antidiscrimination provision preserves the law of any political subdivision of a state. The leave regulation preserves state or local law that provides greater rights. The local layer matters because some of the most significant subnational labor protections of the modern era began as city ordinances: local minimum wages above the state rate, local paid sick days requirements, local fair scheduling rules. Each of these rests on the same savings clauses that protect state statutes, and each is subject to the same one way ratchet.

The local layer also introduces a complication the two sovereign model obscures: the subnational government can preempt its own localities even where the federal government has not preempted the subnational government. A state legislature that dislikes local minimum wage ordinances may forbid its cities from enacting them, and the federal savings clause offers the city no protection against its own state. The federal floor sets the minimum for the nation; the state sets the minimum for its localities where it chooses to; and the city operates in whatever space remains. The worker’s rights in a city with a local wage ordinance therefore depend on three levels of law, not two, and the federal statute’s contribution is still the same floor it always was. The comparison’s three arrangements describe the federal to state relationship; the state to local relationship follows each state’s own constitution and statutes, which this article does not address beyond noting that the federal savings clauses leave it entirely alone.

The Five Errors, Corrected in Detail

The brief for this article names three recurring errors, and the framework corrects two more that surface constantly in practice. Each error comes from reading the wrong provision first, and each dissolves once the arrangement is identified.

The first error is assuming federal law guarantees breaks or paid leave. The assumption usually rests on a vague sense that worker protections of this kind must exist somewhere in federal law, combined with the fact that many workers do receive breaks and paid leave. The correction is that the workers who receive them receive them from jurisdictional law or employer policy, not from any federal mandate. The federal break rule is a pay rule for breaks offered, plus the narrow nursing-mother provision. The federal sick leave rule is nonexistent at the reference date. The federal family leave rule is twelve unpaid weeks. A reader who checks the arrangement finds gaps in all three places, and the gap explains both why some workers have these protections and why others do not: the jurisdiction decided.

The second error is assuming state law is always preempted. The assumption usually comes from encountering ERISA’s broad preemption or the NLRA doctrines and generalizing them to the whole field. The correction is that preemption is selective and each statute names its own selection. Four major statutes expressly preserve more protective jurisdictional law, and section 14(b) expressly authorizes jurisdictional action in the less protective direction. The reader who assumes total preemption will wrongly conclude that a higher jurisdictional minimum wage is void, that a broader jurisdictional discrimination law is displaced, or that a jurisdictional paid leave program is forbidden. Every one of those conclusions is backwards. The savings clauses exist precisely to preserve those laws, and the correct default assumption is the opposite of the error: unless the statute contains a preemption provision or an implied preemption doctrine applies, the jurisdiction may act.

The third error is assuming the minimum wage is the same everywhere. The assumption treats the federal rate as the rule rather than the floor. The correction is section 218: the federal rate governs only where no higher law speaks, and every higher state or municipal rate survives it. The variation across jurisdictions is the statute working as designed, not a patchwork of noncompliance. The deeper correction is methodological. The minimum wage question is the simplest application of the three-step test, and a reader who runs the test on wages will have the habit needed for the harder questions: identify the statute, read the federalism provision, apply the consequence.

The fourth error is assuming that a ceiling means the jurisdiction is powerless on the whole subject. The assumption confuses the preempted slice with the entire field. ERISA preempts jurisdictional laws relating to covered benefit plans, but it does not preempt jurisdictional wage, safety, discrimination, or leave laws, which rest on their own statutes with their own savings clauses. Garmon and Machinists preempt jurisdictional regulation of bargaining conduct, but they do not preempt the jurisdiction’s substantive discrimination law or its wage floor. The ceiling is always bounded by the statute or doctrine that creates it, and the reader must ask what exactly is preempted rather than treating preemption as a general jurisdictional disability. The deemer clause illustrates the boundary-drawing at its finest: the savings clause preserves insurance regulation, the deemer clause prevents definitional capture, and the line between them is where real cases are fought.

The fifth error is assuming that the absence of a federal rule means the absence of any rule. The assumption treats the gap as a void. The correction is that the gap is full of jurisdictional law, varying in content and vigor across the country. Meal breaks, paid sick leave, paid family leave, and dismissal protections exist in many jurisdictions as enacted rights with real enforcement. The gap means there is no federal backstop, not that there is nothing. A reader in a jurisdiction with strong protections on a gap subject has strong protections; the framework’s point is only that those protections come from the jurisdiction and can differ or disappear across the border. The error matters because it leads readers to stop researching at the federal level, when the federal level is exactly where the answer is not.

What is the single most common mistake about workplace rights?

Assuming the federal government guarantees protections it never enacted. Breaks, paid sick leave, paid family leave, and just-cause dismissal protection have no general federal mandate, so workers who have them have them from jurisdictional law. Checking the arrangement first prevents the mistake.

Reading the Law at a Fixed Date

Every statement in this article is anchored to May 1, 2015, and the anchoring is part of the method, not a footnote to it. Employment law has two kinds of facts, and they age differently. The first kind is structural: the text of a savings clause, the three-clause design of a preemption provision, the holding of a Supreme Court decision. These facts are load-bearing and permanent. Section 218 has preserved higher wage laws since its enactment, section 514 has displaced plan-related jurisdictional laws since 1975, and Garmon and Machinists have allocated bargaining questions since 1959 and 1976. A reader in any year can rely on the arrangement column of the table, because Congress writes arrangements to last.

The second kind of fact is numerical, and numbers go stale. The count of approved safety plans changes as jurisdictions enter and leave the program. The minimum wage steps upward in dated increments, five dollars and eighty five cents in 2007, six dollars and fifty five cents in 2008, seven dollars and twenty five cents in 2009, each tied to its effective date. Bills that would create new rights, like the Healthy Families Act of 2015, either advance or die, and their status on any given date is a fact about that date. This article states dated figures with their dates and declines to state figures that change without a date, which is why the safety plan count carries its 2015 sourcing and its August 2015 qualification, and why no jurisdiction-by-jurisdiction wage figures appear.

The date-wall principle follows from the distinction. Developments after the reference date are not part of the law this article describes, and where a later development is relevant enough to mention, it is flagged explicitly as later. The executive order on contractor sick leave, signed September 7, 2015, is the example worked through in the gaps section: it postdates the reference frame, it covers only contract work for the national government, and it is presented with both qualifications rather than folded silently into the analysis. The principle protects the reader in both directions. It prevents the article from claiming as current law something enacted after its date, and it prevents the reader from dismissing the structural analysis because a number has moved. The arrangements endure; the figures are snapshots.

There is a final reason the dating matters for the framework itself. The floor, ceiling, and gap categories are claims about statutory design, and statutory design is visible only when the reader knows which version of the law is being described. A savings clause added by amendment, a preemption doctrine announced in a decision year, a gap narrowed by a targeted provision like the 2010 nursing-mother rule: each of these is a dated event, and the framework’s verdict on a protection can change only when such an event occurs. No event between the enactments described here and May 1, 2015 moved any protection from one column of the table to another. That stability is the finding, and the dates are the evidence for it.

Why does this article avoid current jurisdiction-by-jurisdiction figures?

Because figures change and arrangements do not. The article states dated numbers with their dates, flags later developments at the date-wall, and keeps the table focused on the durable mechanism: which arrangement governs each protection. The mechanism is what answers the practical question in any year.

Deciding Which Arrangement Applies

The series thesis thread for this article is a comparison resolved with a verdict and a named deciding factor. The verdict is that knowing the arrangement beats reading the statute. The deciding factor is the savings or preemption language Congress wrote, because that language, not the substantive rule, determines how much room the jurisdiction has. A reader facing a practical problem can run the same three-step test this article ran.

First, identify the federal statute that touches the subject, if any. Wage questions go to the Fair Labor Standards Act. Safety questions go to the OSH Act. Discrimination questions go to Title VII. Leave questions go to the FMLA. Benefits questions go to ERISA. Union and bargaining questions go to the NLRA. If no federal statute touches the subject at all, the answer is the gap and the inquiry ends in the jurisdiction’s own law.

Second, read the savings or preemption provision of that statute before reading the substantive rule. The savings clauses at 29 U.S.C. 218, 29 U.S.C. 667, 42 U.S.C. 2000e-7 and 29 U.S.C. 2654 announce floors. Section 514 of ERISA at 29 U.S.C. 1144 announces a ceiling. The Supreme Court’s Garmon and Machinists decisions announce ceilings over labor relations conduct. Section 14(b) at 29 U.S.C. 164(b) announces an express state authorization. The provision that controls the federalism question is rarely the provision that states the worker’s right, which is why readers who start with the substantive rule so often misread the jurisdictional answer.

Third, apply the consequence for the jurisdiction. A floor means the worker gets the better of the federal rule and any stronger jurisdictional rule. A ceiling means the worker gets the federal rule and the jurisdiction cannot add to it. A gap means the worker gets whatever the jurisdiction enacted, and in a jurisdiction with nothing additional, the answer is nothing.

This decision procedure also resolves the counter-reading that federal law is the main source of workplace rights. For the protections organized around floors and ceilings, the federal statute is indeed doing heavy work. But for a large share of the protections workers actually invoke, breaks, paid sick time, paid leave and dismissal protection, the federal contribution is silence, and the accurate description is that jurisdictional law does the majority of the work in the areas people ask about most. The assumption that the federal government guarantees these things is the most common error in the field, and it collapses once the reader checks which arrangement applies.

The Savings Clauses Side by Side

The four floors look alike from a distance, but their savings language differs in ways that matter for close questions. Reading the clauses side by side trains the eye this article wants to train: the eye that reads the federalism provision before the substantive rule.

The wage clause, 29 U.S.C. 218(a), is written as a non-excuse provision. It states that no provision of the chapter shall excuse noncompliance with a higher wage, shorter workweek, or stronger child labor law. The drafting technique is to deny the federal Act any shielding effect: an employer cannot point to federal compliance as a defense against a higher jurisdictional standard. The clause names three subjects and stops, which is why its reach is bounded by those subjects. It also names three levels of government, federal, state, and municipal, which is why local wage ordinances share the clause’s protection.

The safety clause, 29 U.S.C. 667, is written as a jurisdictional allocation rather than a non-excuse provision. Subsection (a) preserves state authority where no federal standard exists. Subsections (b) and (c) create an approval regime where federal standards do exist, conditioned on the at least as effective test. The technique here is procedural: instead of simply yielding to higher law, the statute builds a gate through which the jurisdiction must pass, demonstrating agency capacity, inspection rights, personnel, and funding. The safety floor is therefore the most demanding of the four, because the jurisdiction must prove it can hold the floor before it is allowed to stand on it.

The discrimination clause, 42 U.S.C. 2000e-7, is written as a non-exemption provision with a carve-back. It states that nothing in the subchapter shall exempt or relieve any person from liability under present or future state or local law, other than a law purporting to require or permit what the subchapter makes unlawful. Two features stand out. The phrase present or future law makes the savings forward-looking: jurisdictions may legislate after 1964 and still claim the clause’s protection, which is why the added protected characteristics accumulated over decades all rest on the same 1964 language. The carve-back for laws requiring or permitting unlawful acts is the one reservation in the four floors, and it runs only against downward movement into authorized discrimination.

The leave clause, 29 U.S.C. 2654, is written as a two-directional ratchet. Subsection (a) preserves greater rights in contracts and benefit plans; subsection (b) forbids diminishing the federal right by contract or plan; and the regulation extends the upward preservation to state and local law. The technique is the most complete because it binds private ordering as well as public law: neither an employer, nor a union, nor a jurisdiction may go below the federal line, while all three may go above it. No other floor in this article constrains private contracts so explicitly.

The comparison yields a practical rule of thumb. When the question is whether a jurisdiction’s higher standard survives, find the clause and read its technique. A non-excuse clause means the higher standard applies automatically. An approval regime means the jurisdiction must first qualify. A non-exemption clause with a carve-back means the higher standard applies unless it authorizes what federal law forbids. A ratchet means the higher standard applies and the lower one is void even between private parties. The arrangement is the same, a floor, but the mechanism differs, and the mechanism decides the close cases.

Do the savings clauses use identical language?

No. The wage clause denies the federal Act any shielding effect, the safety clause builds an approval gate with an at least as effective test, the discrimination clause preserves present and future law with one carve-back, and the leave clause ratchets in both directions. Same arrangement, four different mechanisms.

Why the Gaps Exist

A reader who has absorbed the floors and ceilings will ask the natural follow-up: why did Congress leave breaks, paid sick time, paid family leave, and just-cause protection out of the federal code? The gaps are not drafting oversights. Each has a distinct explanation, and the explanations show why the gap category is stable rather than temporary.

The break gap reflects the wage statute’s division of labor between pay rules and provision rules. When Congress wrote the Fair Labor Standards Act, it regulated the price of labor, the minimum wage and the overtime premium, not the scheduling of the workday. The Department of Labor’s regulations filled the interstice with pay consequences: short breaks must be paid if offered, meal periods need not be. Requiring the breaks themselves would have moved the statute from price regulation into time management, a step Congress never took in the general case. The nursing-mother provision of 2010 is the exception that confirms the pattern: Congress acted when a specific, sympathetic, and narrowly defined need arose, and it wrote the provision with tight limits, unpaid breaks, non-exempt employees only, a one-year window, and a small-employer hardship exemption. The general gap survived because the general question was never legislated.

The paid sick leave gap reflects a legislative near-miss rather than a considered omission. The Healthy Families Act of 2015 would have created a national standard, and its failure to advance left the field to the jurisdictions. The gap is therefore contingent in a way the break gap is not: it persists because a bill did not pass, not because Congress decided the subject belonged to the jurisdictions as a matter of principle. That contingency is worth stating plainly, because it affects how a reader should think about the gap’s durability. Some gaps are structural, and some are merely the current score.

The paid family leave gap reflects the political economy of the 1993 compromise. The FMLA passed as an unpaid-leave statute because unpaid leave was the version that could assemble a majority: it imposed no direct wage cost on employers and no new payroll tax. Paid leave would have required a funding mechanism, either employer mandates or a social insurance tax, and neither commanded the necessary coalition. The savings clause at section 2654 then did exactly what savings clauses do: it let the jurisdictions that wanted paid programs build them on the federal floor, producing the current distribution in which the federal government guarantees time and some jurisdictions guarantee pay. The gap is the product of a compromise that bought passage with unpaid leave and exported the pay question to the jurisdictions.

The just-cause gap reflects the deepest root of all: the common law. Employment at will is not a federal creation but an inherited default of American employment law, older than any of the statutes in this article. Congress has never enacted a general just-cause requirement because doing so would have displaced the default across the entire economy, a step far more intrusive than any of the targeted prohibitions on discriminatory or retaliatory discharge that Congress did enact. The federal constraints on firing are therefore a patchwork of specific forbidden reasons rather than a general required reason, and the patchwork leaves the at-will default standing everywhere it has not been pierced. Montana’s general just-cause statute is the exception that proves how firmly the default holds elsewhere: one jurisdiction in fifty displaced it, and the other forty-nine did not.

Where the Arrangements Overlap

Real workplace questions rarely arrive in single-statute packaging, and the framework must handle subjects where two arrangements meet. Three overlaps recur often enough to deserve explicit treatment.

The first is leave taken under a benefit plan. A worker’s medical leave implicates the FMLA, a floor, and the employer’s health plan, a ceiling. The FMLA question is whether the jurisdiction added paid leave or longer leave on top of the federal twelve unpaid weeks. The benefits question is whether the jurisdiction can mandate what the health plan covers during that leave, and ERISA section 514 answers that the mandate is displaced insofar as it relates to the plan. The worker can therefore hold a jurisdictional paid-leave right and a federally governed health plan at the same time, with different levels of government controlling each. The overlap resolves cleanly once each question is assigned to its statute: the floor governs the leave, the ceiling governs the plan, and neither answer contaminates the other.

The second is safety regulation touching workers’ compensation. Workplace safety standards are a floor under the OSH Act’s state plan mechanism, while workers’ compensation is a jurisdictional system the federal safety statute does not preempt. A state plan that sets standards at least as effective as the federal ones operates alongside the jurisdiction’s compensation system without conflict, because the two address different questions: the plan prevents injuries, the compensation system pays for them. The overlap is cooperative rather than competitive, which is why it generates less litigation than the benefits overlap. The framework’s contribution here is negative but useful: it tells the reader that no preemption question arises, because neither statute claims the other’s territory.

The third is discrimination in a unionized workplace. Title VII’s floor means the jurisdiction may add protected characteristics and reach smaller employers. The NLRA’s ceiling means the jurisdiction may not regulate the bargaining conduct through which a union pursues those protections. A jurisdiction can therefore expand what counts as discrimination while being unable to dictate how unions and employers bargain over anti-discrimination provisions in their contracts. The floor and the ceiling operate on different planes, one substantive and one procedural, and the worker’s rights are the sum of both: broader substantive protection under the jurisdictional add-on, with bargaining conduct governed by the federal framework both Garmon and Machinists describe.

The general lesson of the overlaps is that the arrangement attaches to the statute, not to the workplace. A single workplace is governed by floors, ceilings, and gaps simultaneously, and the reader’s task is to sort each question into its statute before applying the test. The three-step procedure from the decision section handles overlaps without modification: identify the federal statute for each question separately, read each statute’s federalism provision, and apply each consequence. Confusion arises only when the reader tries to assign one arrangement to the whole workplace instead of one arrangement to each question.

The Enforcement Map

The arrangement also decides who enforces the law, which is the framework’s most practical payoff after the substantive answer. Floors are enforced by two sovereigns at once. The Department of Labor’s Wage and Hour Division enforces the federal wage floor while jurisdictional agencies enforce higher local rates. Federal OSHA enforces federal standards directly in jurisdictions without approved plans and monitors the plans where they exist. The Equal Employment Opportunity Commission enforces the federal title while deferring charges to jurisdictional fair employment agencies under worksharing arrangements. The Department of Labor enforces the FMLA floor while jurisdictional agencies administer paid leave programs built on it. In every floor case the worker may have two enforcers, and the practical advice that follows from the structure, without individualized guidance, is that the forum follows the claim: the federal agency for the federal floor, the jurisdictional agency for the jurisdictional addition.

Ceilings are enforced in federal forums. ERISA’s preemption means benefit disputes are litigated under federal standards, and the Department of Labor’s Employee Benefits Security Administration polices plan conduct. The NLRA’s implied preemption means bargaining conduct disputes go to the National Labor Relations Board, not to state courts or agencies. The uniformity the ceiling produces is therefore institutional as well as substantive: one set of enforcers applying one body of law. A worker or employer who takes a ceiling question to a jurisdictional forum will be redirected, which is Garmon’s primary jurisdiction point restated as practical advice.

Gaps are enforced, if at all, by the jurisdiction alone. A break claim, a paid sick leave claim, a paid family leave claim beyond the federal unpaid entitlement, or a just-cause claim exists only where the jurisdiction created it, and only the jurisdiction’s agency or courts can hear it. There is no federal backstop. This is the enforcement dimension of the surprise the gap category produces: not only is there no federal right, there is no federal forum either. The worker in a jurisdiction with no additional law has nowhere to take the claim, because there is no claim.

The enforcement map completes the One Test for this article. A reader who can name the arrangement for a protection, state the savings or preemption provision that creates it, identify the consequence for a jurisdiction with no additional law, and say which sovereign enforces the result has mastered the organizing principle. Everything else in employment law is detail hung on that frame.

Uniformity and Experimentation, Stated With Equal Care

Preemption and jurisdictional variation are contested on both federalism and worker protection grounds, and the neutrality of this article requires presenting each position by its statutory source and with equal care in attribution. The uniformity argument holds that a single national standard for benefit plans and labor relations prevents a fragmented compliance landscape, lets multi-jurisdiction employers operate under one rule, and protects the federal interest in an integrated national economy; its proponents point to ERISA section 514 and the NLRA preemption doctrines as deliberate congressional choices to secure that uniformity. The experimentation argument holds that the floors Congress wrote for wages, safety, discrimination and leave are invitations to tailor protections to local economic conditions, cost of living and policy preferences, and that variation across jurisdictions is the designed product of the savings clauses; its proponents point to 29 U.S.C. 218, 29 U.S.C. 667, 42 U.S.C. 2000e-7 and 29 U.S.C. 2654 as the textual evidence. This article characterizes no jurisdiction’s choices and gives no individualized guidance; its contribution is the structural map on which any reader can locate the argument that applies to the protection at hand.

The uniformity position draws its strongest evidence from the benefits field. Defenders of ERISA section 514 argue that employers operating retirement and health plans across dozens of jurisdictions cannot administer those plans under fifty different sets of mandates, and that the relate to standard, broad as it is, purchases something real: a single body of plan law that lets a multi-jurisdiction employer promise the same benefits to every worker. The argument extends to labor relations, where defenders of the Garmon and Machinists doctrines contend that a national labor policy requires a single tribunal and a single balance of economic weapons, and that jurisdictional regulation of bargaining conduct would fragment the policy Congress assigned to the Board. On this view, the ceiling is not a restriction on the jurisdictions but a precondition of a functioning national system, and the jurisdictions retain ample room where Congress left them room, which is everywhere the floors and the section 14(b) authorization operate.

The experimentation position draws its strongest evidence from the floors. Defenders of jurisdictional variation argue that the savings clauses are invitations to tailor protections to local conditions, and that the resulting diversity is a designed feature rather than a failure of coordination. A higher minimum where the cost of living is higher, broader discrimination coverage where the political consensus supports it, paid leave where the fiscal capacity exists: each is presented as a jurisdiction solving its own problem with tools Congress deliberately left in jurisdictional hands. The classic formulation holds that jurisdictions serve as laboratories, testing policies at small scale before they are considered nationally, and proponents point to the paid leave programs built on the FMLA floor as the mechanism working exactly as intended. On this view, uniformity would be the real cost, because it would freeze protections at the level of the least willing jurisdiction and forbid the willing from moving first.

The gap category complicates both positions, which is why neutrality requires stating the complication. Defenders of uniformity note that the gaps produce the widest variation of all, with workers in some jurisdictions holding rights that workers elsewhere lack entirely, and they present this as evidence that leaving subjects to the jurisdictions guarantees inequality of protection. Defenders of experimentation reply that the gaps show what happens without any congressional baseline, and that the answer is more floors rather than fewer jurisdictions: enact a national minimum and let the jurisdictions build, as the wage, safety, discrimination, and leave statutes did. The disagreement is genuine, it is longstanding, and it maps directly onto the arrangements this article describes. The reader who understands floors, ceilings, and gaps understands the structure of the federalism debate itself, because the debate is conducted in those three categories whether or not the participants use the names.

What to Study First

This article carries the study-path recommendation, which is why both companion tools apply. The sequence that serves a reader with a practical problem runs from the arrangement to the statute: first learn the floor, ceiling and gap framework, then learn the savings or preemption provision of the statute that governs the question, and only then read the substantive rule. A reader who reverses that order, reading the substantive rule first, will know what the federal standard says and will have no idea whether the jurisdiction can change it, which is the question that actually matters. The series study guide places this article in the larger sequence for readers deciding what to study first, and the reasoning drills that reinforce the three-step decision test live in a legislation study notebook alongside a civics study tool for the constitutional background.

The study path can be sequenced concretely. Begin with the four savings clauses, reading each in full: 29 U.S.C. 218, 29 U.S.C. 667, 42 U.S.C. 2000e-7, and 29 U.S.C. 2654. Note the drafting technique of each, the non-excuse provision, the approval gate, the non-exemption with carve-back, and the two-directional ratchet, because the technique decides the close cases. Next read the ceiling provisions: ERISA section 514 with its three clauses, then the one-paragraph holdings of Garmon and Machinists, distinguishing the forum rationale of the former from the substantive rationale of the latter. Then read section 14(b) as the special case, noting the difference between its literal text on membership agreements and its practical role as the basis for right-to-work laws. Then study the gaps, not by reading statutes but by confirming absences: the Department of Labor’s statement that no general break requirement exists, the unpaid character of FMLA leave, the at-will default with its patchwork of specific forbidden reasons for discharge. Finally, run the worked examples until the three-step test is reflexive: identify the statute, read the federalism provision, apply the consequence.

That sequence is the study-path recommendation this article carries, and it is why both companion tools apply. The framework first, the provisions second, the substance third, and the test until it is habit. A reader who studies in that order will rarely need to look up the answer to a new question, because the arrangement will supply it.

A Last Word on the Deciding Factor

The series thesis thread for this article promised a comparison resolved with a verdict and a named deciding factor, and the deciding factor deserves a final, explicit statement. It is the savings or preemption language Congress wrote into each statute. Not the substantive right, not the agency’s gloss, not the weight of commentary, but the clause that allocates authority between the two sovereigns. Section 218 decides wage questions. Section 18 of the safety statute decides safety questions. Section 708 decides discrimination questions. Section 404 decides leave questions. Section 514 decides benefits questions. Garmon and Machinists decide bargaining conduct questions. Section 14(b) decides union security questions. And where no clause exists because no statute exists, the gap decides, which is itself a decision: the jurisdiction governs alone.

This is why the article insists on reading the federalism provision before the substantive rule. The substantive rule tells the reader what the national standard requires. The federalism provision tells the reader whether the jurisdiction may change it, and in which direction, and through what procedure. Every practical question about workplace rights is really the second question wearing the first question’s clothes. A worker asking about the minimum wage is asking whether the jurisdiction raised the floor. A worker asking about paid leave is asking whether the jurisdiction built on it. A worker asking about breaks is asking whether the jurisdiction filled the silence. Learn to hear the federalism question inside the practical question, and the statute book opens in the right order every time.

The verdict the comparison demands is simple. American employment law is not a federal system with jurisdictional supplements but three arrangements operating at once, floors that invite more protection, ceilings that forbid it, and gaps where the federal government never spoke. The deciding factor is always the savings or preemption language Congress wrote. Learn to read that language first, and most practical questions about workplace rights answer themselves before the substantive rule is ever opened.

Five worked examples show the three-step test in motion. Each begins with the kind of question a worker, an employer, or a student actually asks, then runs the test: identify the federal statute, read its savings or preemption provision, and apply the consequence for the jurisdiction.

The first question is whether a worker must be paid the federal minimum or something higher. The federal statute is the Fair Labor Standards Act. Its savings provision is section 18, 29 U.S.C. 218, which preserves any higher state or municipal minimum. The arrangement is a floor. The consequence is that the worker receives the highest applicable rate: the local ordinance if one exists, otherwise the state rate if one exists, otherwise the federal seven dollars and twenty five cents per hour. The test also disposes of the follow-up question, whether a jurisdiction could set a lower rate. The floor runs only upward, so the answer is no, and the federal rate binds every covered employer where the jurisdiction has nothing higher.

The second question is whether an employer must provide paid sick leave. The search for a federal statute comes up empty for the general case: no federal statute required paid sick leave on May 1, 2015, the temporary and contractor-only measures belonging to other dates and other scopes. The arrangement is a gap. The consequence is that the worker’s right is whatever the jurisdiction enacted, and in a jurisdiction with no paid sick leave law the worker has no such right under federal law. The test exposes the common error directly: the assumption that a federal guarantee exists collapses at step one, because there is no federal statute to read.

The third question is whether a jurisdiction can require employee benefit plans to cover a particular treatment. The federal statute is ERISA. Its preemption provision is section 514, 29 U.S.C. 1144, whose relate to standard supersedes state laws relating to covered plans, subject to the insurance savings and deemer clauses. The arrangement is a ceiling. The consequence is that the mandate is displaced insofar as it relates to covered plans; the jurisdiction’s remaining path runs only through genuine insurance regulation of insured arrangements, and the deemer clause blocks the definitional shortcut of deeming self-insured plans to be insurers. The worker’s benefit rights are therefore the federal rule, and the jurisdiction cannot add a mandate on top of it.

The fourth question is whether a union can require new hires to pay fees as a condition of employment. The federal statute is the National Labor Relations Act, which as amended permits union security agreements. The controlling provision is section 14(b), 29 U.S.C. 164(b), which expressly authorizes jurisdictions to prohibit agreements requiring membership as a condition of employment. The arrangement is the express downward authorization. The consequence depends entirely on the jurisdiction: where no prohibiting law exists, the security agreement stands; where the jurisdiction enacted one, the agreement is void and support must be voluntary. The same federal statute thus yields opposite answers in different jurisdictions, by congressional design.

The fifth question is whether a worker at a very small business is protected against discrimination. The federal statute is Title VII, which reaches employers above a statutory threshold. Its savings provision is section 708, 42 U.S.C. 2000e-7, which preserves state and local fair employment laws except those requiring or permitting what the title forbids. The arrangement is a floor. The consequence is that the worker’s protection depends on whether the jurisdiction extended coverage to smaller employers: where it did, the broader jurisdictional rule governs; where it did not, the federal threshold leaves the worker outside the title. The test shows why the answer cannot be given from the federal statute alone.

Which arrangement produces the most uniform national rules?

The ceiling. ERISA section 514 displaces state laws relating to covered benefit plans, and the Garmon and Machinists doctrines displace state regulation of core bargaining conduct. Floors produce variation by design, and gaps produce the widest variation of all, since the jurisdiction’s law is the entire field.

Frequently Asked Questions

Q: Do state labor protections beat federal ones?

They do when the federal statute sets a floor rather than a ceiling, which is the most common arrangement. Four major statutes contain express savings language preserving more protective state law: the Fair Labor Standards Act at 29 U.S.C. 218, the OSH Act at 29 U.S.C. 667, Title VII at 42 U.S.C. 2000e-7, and the FMLA at 29 U.S.C. 2654. Where a state has enacted a higher minimum wage, a broader anti-discrimination rule, or a paid leave program, the state rule governs because the federal provision expressly yields to it. State protections do not beat federal ones where Congress built a ceiling, most notably ERISA section 514, which supersedes state laws relating to employee benefit plans, or where the Supreme Court’s Garmon and Machinists doctrines impliedly preempt state regulation of labor relations conduct. And where federal law is silent, on meal breaks, paid sick leave, paid family leave, or just-cause dismissal protection, there is no federal rule to beat at all: state law is the entire body of applicable law. The practical method is to identify the arrangement first, then read the statute.

Q: Is federal labor law a floor or a ceiling?

It is both, plus a third category: silence. Most federal labor and employment statutes are floors. The wage and hour statute, the workplace safety statute, the civil rights employment title, and the leave statute each expressly preserve more protective state law, so states may set higher minimum wages, run approved safety programs, bar discrimination by smaller employers or on additional grounds, and provide longer or paid leave. A small number of provisions are ceilings. ERISA section 514 preempts state laws relating to employee benefit plans with only narrow exceptions, and the Supreme Court’s Garmon and Machinists decisions preempt state regulation of conduct that the National Labor Relations Act protects, prohibits, or deliberately leaves to the free play of economic forces. And on several subjects workers care about most, meal and rest breaks, paid sick leave, paid family leave, and just-cause protection, federal law is silent, so the arrangement is neither floor nor ceiling but a gap filled entirely by state law. Knowing which of the three applies to a given protection answers most practical questions faster than reading any single statute.

Q: Does federal labor law require meal breaks?

No, with one narrow exception. The Wage and Hour Division states plainly that federal law does not require lunch or coffee breaks. Federal regulation addresses only the pay consequences of breaks an employer chooses to offer: short rest periods of five to twenty minutes must be paid if offered, while bona fide meal periods, typically thirty minutes or more with the employee fully relieved of duties, are not compensable. Those are rules about pay for breaks, not rules requiring breaks to exist. The exception is FLSA section 7(r), 29 U.S.C. 207(r), effective March 23, 2010, which requires reasonable unpaid break time for a non-exempt employee to express breast milk for her nursing child for one year after the child’s birth, plus a private non-bathroom space; employers with fewer than fifty employees are exempt if compliance would impose an undue hardship. Outside that provision, whether a worker receives meal or rest breaks depends entirely on the law of the jurisdiction, and in a jurisdiction with no break law the worker has no federal break right at all.

Q: Which states have higher minimum wages than federal labor law?

This article states no jurisdiction-by-jurisdiction figures because they change with each legislative session and would date the analysis. The durable answer is structural. Section 18 of the Fair Labor Standards Act, 29 U.S.C. 218, provides that no provision of the federal law excuses noncompliance with any state law or municipal ordinance establishing a higher minimum wage, and the Congressional Research Service confirms that where a state enacts a more protective minimum wage law, the state law applies. The federal rate of seven dollars and twenty five cents per hour, set by the Fair Minimum Wage Act of 2007 and effective July 24, 2009, therefore operates only as a floor. Many jurisdictions have used the opening that section 218 preserves to set higher rates, and the mechanism guarantees that a worker covered by both always receives the higher of the two. A reader who needs the current figure for a particular jurisdiction should consult the Department of Labor’s published wage data for that jurisdiction rather than any fixed list. The arrangement itself, a federal floor that yields to higher state law, has not changed since the savings clause was enacted.

Q: Does federal labor law preempt state rules?

Sometimes, and the statute or doctrine at issue decides. The broadest express preemption in the field is ERISA section 514, 29 U.S.C. 1144, which supersedes any and all state laws insofar as they relate to covered employee benefit plans, subject to a savings clause for genuine insurance, banking, and securities regulation and a deemer clause preventing states from capturing self-insured plans by deeming them insurers. In labor relations, preemption is implied rather than written: Garmon, 359 U.S. 236 (1959), defers state regulation of conduct arguably protected or prohibited by the National Labor Relations Act to the National Labor Relations Board, and Machinists, 427 U.S. 132 (1976), bars states from regulating conduct Congress left to the free play of economic forces. But preemption is not the default. Four major statutes expressly preserve more protective state law, and section 14(b) of the NLRA expressly authorizes states to prohibit union security agreements. The common error is assuming preemption is total; the accurate statement is that it is selective, and each statute’s own savings or preemption language names the selection.

Q: Is at will employment required by federal labor law?

Federal law does not require at-will employment, but it preserves it by never displacing it. No federal statute imposes a general just-cause requirement on private employers. The federal constraints on discharge are prohibitions on specific illegal reasons for firing, discrimination barred by Title VII, retaliation barred by whistleblower and rights provisions, and similar targeted rules, not a requirement that the employer prove cause before terminating. The United States default is employment at will: most workers may be terminated for any reason that is not independently unlawful, a doctrine established by the at-will default plus the absence of any federal just-cause statute. Montana is the one jurisdiction with a general just-cause requirement after a probationary period under its Wrongful Discharge from Employment Act. Elsewhere, protection against being fired without cause exists only if the jurisdiction created it, and federal law neither creates it nor prevents the jurisdiction from doing so. A reader who assumes the federal government guarantees protection against arbitrary dismissal is relying on a guarantee that was never enacted.

Q: What is right to work under federal labor law?

Right to work is the name given to state laws authorized by section 14(b) of the National Labor Relations Act, 29 U.S.C. 164(b). That provision states that nothing in the Act shall be construed as authorizing agreements requiring membership in a labor organization as a condition of employment in any state or territory where such agreements are prohibited by state or territorial law. In practice, the state laws built on this authority prohibit compelling workers to join a union or pay fees, including agency-fee arrangements, as a condition of keeping their jobs. In jurisdictions without such laws, unions may negotiate security agreements requiring financial support from the workers they represent; in jurisdictions with them, those agreements are void and the union represents the bargaining unit while collecting support only from volunteers. Section 14(b) is the clearest instance in this series of a federal statute authorizing states to move in the less protective direction, and it is why one national labor statute produces two very different regimes. The provision invites states downward where the floor statutes invite them upward.

Q: Should a student study federal or state labor law first?

Start with the federal arrangement, then read the state. The three-step method this article recommends is: identify the federal statute that touches the subject, read its savings or preemption provision before the substantive rule, and apply the consequence for the jurisdiction. A student who reads the substantive rule first will know what the federal standard says but not whether the jurisdiction can change it, which is the question that actually matters in practice. The floor, ceiling, and gap framework gives the student a single organizing principle: most federal statutes set floors the state may exceed, a small number preempt state law entirely, and on several subjects workers ask about most there is no federal rule at all. Once the arrangement is identified, the state inquiry is bounded and efficient: for a floor, check whether the jurisdiction added anything; for a ceiling, stop; for a gap, the jurisdiction’s law is the whole field. The series study guide places this article in the larger sequence for readers deciding what to study first.

Q: What does ERISA preemption mean for state benefit mandates?

It means a state generally cannot mandate the content of employee benefit plans. ERISA section 514, 29 U.S.C. 1144, supersedes any and all state laws insofar as they relate to covered employee benefit plans, and courts have read “relate to” to reach laws with a connection to or reference to plans, not only laws that regulate plans directly. Two refinements matter. The savings clause preserves state laws that regulate insurance, banking, or securities, so an insurance mandate directed at insurers can survive. The deemer clause then prevents a plan or its trust from being deemed an insurer for purposes of state regulation, which stops states from reaching self funded plans through the insurance door. The net effect is that benefit plan content is a national subject governed by federal law and the plan terms.

Q: How does a state get an approved OSHA safety plan?

Section 18 of the OSH Act, 29 U.S.C. 667, sets a multi-part approval test. The state submits a plan to the Secretary of Labor showing that it designates a responsible state agency, provides standards and enforcement at least as effective as the federal standards for the same issues without unduly burdening interstate commerce, provides entry and inspection rights at least as effective as section 8 with no advance notice, demonstrates legal authority and qualified personnel, and provides adequate funding. The phrase at least as effective is the operative standard: the state program must match or exceed the federal one, never fall below it. A 2015 joint letter from the National Institute for Occupational Safety and Health and the Occupational Safety and Health Administration reported 28 approved state plans that year, a figure that reached that number with a Massachusetts public-employee plan approved in August 2015, so the count is stated here as a dated 2015 figure rather than as the precise count on May 1, 2015. Where no federal standard is in effect, subsection 18(a) lets a state assert jurisdiction directly without a plan at all.

Q: What are the Garmon and Machinists preemption doctrines?

They are the two judge made doctrines through which the National Labor Relations Act preempts state regulation of labor relations conduct. Garmon, from San Diego Building Trades Council v. Garmon, 359 U.S. 236 (1959), holds that the Act impliedly preempts state regulation of conduct that is arguably protected by section 7 or arguably prohibited by section 8, deferring jurisdiction to the National Labor Relations Board. The word arguably matters: even possible coverage sends the state aside. Machinists, from Lodge 76, IAM v. Wisconsin Employment Relations Commission, 427 U.S. 132 (1976), holds that the states may not regulate conduct Congress intended to leave unregulated and controlled by the free play of economic forces, such as the peaceful concerted refusal to work overtime during bargaining at issue in that case. Together they leave states almost no independent room in labor relations.

Q: How does Section 14(b) let states ban union security agreements?

Section 14(b) of the National Labor Relations Act, 29 U.S.C. 164(b), provides that nothing in the Act authorizes the execution or application of agreements requiring membership in a labor organization as a condition of employment in any state or territory where state law prohibits them. Under federal law, section 8(a)(3) otherwise permits union security agreements, so without section 14(b) a state ban would collide with the federal permission. Section 14(b) resolves the collision in advance by withdrawing federal authorization wherever the state has acted. In practice it is the statutory basis for state right to work laws, which in most of those states also bar compulsory fee arrangements. The provision is the clearest instance in this series of a federal statute expressly authorizing states to move in a less protective direction than the federal baseline.

Q: What are union security agreements under federal labor law?

They are agreements between an employer and a union, permitted by section 8(a)(3) of the National Labor Relations Act, that require employees to join the union or pay its equivalent as a condition of keeping their jobs. The classic forms are the union shop, where new hires must join within a stated period, and the agency shop, where nonmembers must pay fees covering representation costs. The Supreme Court has limited what those payments may fund, excluding expenditures unrelated to collective bargaining and contract administration. In states with right to work laws enacted under section 14(b), 29 U.S.C. 164(b), such agreements are void because the state has prohibited them and the federal statute withdraws its authorization there. In states without such laws, the agreements remain enforceable as federal law permits.

Q: Can a state prohibit discrimination that federal law does not cover?

Yes. Title VII section 708, 42 U.S.C. 2000e-7, preserves any state law that adds liability rather than subtracting it, which means a state may prohibit employment discrimination on grounds the federal title does not reach. By 2015 several states had done so for categories including sexual orientation and marital status, and those protections survived because they imposed duties beyond the federal floor. The same clause lets states cover smaller employers than the federal fifteen employee threshold, since a broader coverage rule adds protection rather than contradicting the federal one. The only state laws the clause defeats are those that purport to require or permit an act that would be an unlawful employment practice under the federal title. The discrimination floor is therefore exactly federal in some states and substantially higher in others.

Q: Can a state require longer or paid leave beyond the FMLA?

Yes, and several had by 2015. FMLA section 404, 29 U.S.C. 2654, provides that nothing in the Act diminishes rights established by collective bargaining agreements or benefit programs that provide greater family or medical leave, and the implementing regulation extends the principle to state and local law. The federal entitlement is up to twelve weeks of unpaid, job protected leave per year for qualifying employees. Because the statute preserves greater rights, states were free to enact longer leave periods, leave for additional family members, and paid family leave programs financed through payroll contributions, and the federal statute neither required nor blocked them. A worker in a state that enacted such a program therefore holds leave rights the federal statute never created, which is the floor arrangement working exactly as its savings clause designed.

Q: Did federal law require paid sick leave on May 1, 2015?

No. On the reference date of this article, no general federal paid sick leave requirement existed. The Healthy Families Act, introduced in February 2015 as H.R. 932 with Senate companion S. 497, would have created a national paid sick days standard, but it saw no action in the House subcommittee. Executive Order 13706, which established paid sick leave for employees of federal contractors, was signed on September 7, 2015, more than four months after the reference date, and it covers only work on federal contracts, so it was not in force on May 1, 2015 and must not be described as a general federal paid sick leave law. The first general federal paid sick requirement was the temporary Families First Coronavirus Response Act of 2020. On the reference date, paid sick leave was a matter of state law and employer policy alone.

Q: What protections exist where a state has no wrongful discharge law?

Only the targeted federal and state protections against firing for specific forbidden reasons. No general federal just cause statute exists; the at will default governs, so an employer may terminate for any reason that is not independently unlawful. The unlawful reasons are the ones particular statutes name: discrimination under Title VII and the other federal antidiscrimination statutes, retaliation for protected activity under statutes including the National Labor Relations Act and the occupational safety laws, and similar state law protections. Only Montana, through its Wrongful Discharge from Employment Act, supplies a general just cause protection after a probationary period. In the remaining states, a worker fired for a reason that is merely unfair, as opposed to discriminatory or retaliatory, has no legal claim under either sovereign’s law.

Q: What happens when a state law conflicts with a federal floor?

There is no true conflict, because the floor is designed to absorb the difference. Where a federal statute sets a floor through a savings clause, the governing rule is the more protective of the two: the worker receives the higher minimum wage, the broader anti-discrimination protection, the longer or paid leave, or the stronger safety standard. The federal provision yields by its own terms. Section 18 of the Fair Labor Standards Act states that no provision of the Act excuses noncompliance with a higher state standard. Section 708 of Title VII preserves state fair employment laws. Section 404 of the FMLA preserves greater leave rights. The only genuine conflicts arise where the federal provision is a ceiling rather than a floor: a state mandate relating to a covered benefit plan collides with ERISA section 514, and a state regulation of protected or prohibited bargaining conduct collides with Garmon or Machinists preemption. In those cases the federal rule displaces the state one. The first step is therefore always to identify the arrangement before asking which law wins.

Q: How does a reader decide which law answers a workplace question?

Name the protection, then ask what the governing federal statute says about state law. A savings clause, as in 29 U.S.C. 218, 29 U.S.C. 667, 42 U.S.C. 2000e-7, or 29 U.S.C. 2654, signals a floor: the worker gets the better of the federal and state standards. Express preemption language, as in ERISA section 514, 29 U.S.C. 1144, or the Garmon and Machinists doctrines in labor relations, signals a ceiling: the federal rule is the only rule. The absence of any general federal rule, as with meal breaks, paid sick leave, paid family leave, and just cause dismissal, signals a gap: state law is the only law. The arrangement, not the subject matter, determines where to look first.

Q: Why do employment rights differ so sharply across state lines?

Because the three arrangements distribute authority differently, and the gap category makes the jurisdiction the entire field. On floor subjects, rights differ by the amount the jurisdiction added above the federal minimum: higher minimum wages, broader discrimination coverage, paid leave programs. On ceiling subjects, rights are uniform because the jurisdiction cannot add anything: benefit plan rules and core bargaining conduct look the same everywhere. On gap subjects, rights differ completely, because there is no federal rule at all and the jurisdiction’s law is the whole body of applicable law. Meal breaks, paid sick leave, paid family leave, and just-cause dismissal protection exist in some jurisdictions and not in others, with nothing in between. The sharpest differences therefore appear exactly where workers ask most often, which is why the assumption of a uniform national floor of workplace rights fails on contact with the structure. The variation is not a malfunction. It is the product of savings clauses that invite it, preemption provisions that forbid it in narrow fields, and silences that leave entire subjects to the jurisdictions.