Employment Law for Employers

Employment is the practice area where an ordinary decision, made quickly and in good faith, turns into a claim two years later. Nobody sets out to misclassify a contractor or sign an unenforceable covenant. It happens because the paperwork was inherited, or borrowed from a friend’s company, or written for a business in a different state at a different size — and then never looked at again while the company grew around it.

Connecticut, New York and Massachusetts each apply their own test to whether a worker is an employee or a contractor, their own rules on when a final paycheck is due, and their own view of how far a restrictive covenant may reach. A handbook that is correct in one of them can be wrong in the next. That is not an argument against hiring across state lines. It is an argument for knowing which state’s rules apply before the offer goes out.

Where it usually goes wrong: an offer letter that promised something the handbook contradicts. A contractor who is functionally an employee. A non-compete drafted for the wrong state. Work product nobody assigned to the company. A separation handled on a phone call. Those are the same questions a business legal audit asks across the whole employment file at once, rather than one hire at a time.

The moment to set expectations, and to protect the company’s IP, is before someone starts.

A short at-will offer letter covers most hires: role, pay, start date, who they report to, and that employment is at will. Just as important is what it should not say — an annual salary framed as a yearly guarantee, or a promise about equity the equity documents will not match. A fuller employment agreement makes sense once equity, restrictive covenants, or negotiated role terms are involved. The right document depends on the hire.

Equity and option grant paperwork gets coordinated with the cap table rather than described in an interview and reconciled later.

Onboarding is where confidentiality and invention-assignment terms get signed — before the first day, as a condition of the offer. Asking someone to sign an invention assignment after they have started is a request they can decline, and in some states an agreement signed later without fresh consideration stands on weaker ground. That single timing decision is the most common reason a company does not own what it paid to build.

Deliverables: an offer letter that says what you meant, including what it does not promise · at-will language that survives contact with the handbook · confidentiality and invention-assignment terms signed at the start · equity or bonus terms written down.

Also handled: policy acknowledgments · multi-state offer paperwork where the hire sits outside CT, NY, or MA.

Nobody decides to misclassify anyone. It happens one reasonable decision at a time. The first person comes on as a contractor because payroll and benefits are a commitment you are not ready to make. Eighteen months later there are five of them, they are on your Slack, they attend standup, and one has not worked for anyone else since the day they started. Classification does not care what you intended at the start. It looks at how the relationship works now.

Under an ABC-style test the worker is presumed to be an employee, and the business has to prove all three parts, not the best two. A — free from direction and control, in practice as well as on paper. B — outside the usual course of your business. C — independently established in that trade. The middle prong decides most cases: a restaurant hiring a plumber is comfortably outside its usual course; a software company hiring a contract developer is not, and no amount of independence fixes a prong that asks what the work is.

Connecticut and Massachusetts both use ABC-style standards in significant contexts. New York leans on a common-law control analysis. Federal tax uses its own multi-factor approach. The same worker can be assessed differently for unemployment, wage and hour, workers’ compensation, and tax, which is why “are they a contractor?” usually deserves the answer “for which purpose?”

Getting it wrong compounds: unpaid payroll taxes with interest and penalties, overtime and minimum wage exposure, unemployment and workers’ compensation contributions, missed benefits, and in some circumstances personal exposure for the decision-makers. Some wage statutes provide multiplied damages and fee-shifting. And it rarely stays with one person — these start when a contractor is let go and files for unemployment, or is injured, or falls out over an invoice, and the finding then reaches everyone in a similar role.

Deliverables: a classification analysis against the test the relevant state actually applies · contractor agreements consistent with how the work is really performed · a written note on exposure where a role is already misclassified · a reclassification plan where that is the safer answer.

Also handled: multi-state guidance for remote teams · deliberate risk remediation, sequenced so a conversion does not read as an admission.

A handbook is the record of what the company told people, and it is what a defense rests on when somebody says they were never told.

It is usually not legally required as a single document, though specific policies are required in some states as headcount grows. The reason to have one is evidentiary. Without it, every dispute gets decided on he-said-she-said.

The policies that carry weight are leave, complaints, discipline, and technology use, written for the states your people actually work in. Remote and hybrid teams make that harder, not easier — a policy that holds up in one office does not automatically travel.

The most common own goal is language that reads like a promise. Discipline steps described as a sequence, job security implied, benefits stated without qualification. Any of those can be read as a contract. That is what the at-will and disclaimer language is for, and it has to survive the rest of the document rather than sit on page two contradicting page nine.

Deliverables: a handbook written for the states your people actually work in · the policies that carry legal weight · an acknowledgment page that records receipt in a way you can rely on later · a review of the handbook you already have, with the gaps listed.

Also handled: required state policies as headcount crosses statutory thresholds · remote and hybrid policies · policy updates as the law and the company change, which is one of the standing items on an outside general counsel retainer.

There are two kinds of non-compete: the kind that works because it is enforceable, and the kind that works because nobody tests it. The second is more common than anyone admits — a restriction copied from a template, dropped into every offer letter regardless of role, covering the whole country for three years. It deters people who assume it must mean something, and falls apart the first time a departing employee hires a lawyer who reads it closely.

The useful question is not “how much can we restrict?” but “what are we actually protecting, and what is the narrowest restriction that protects it?”

Connecticut applies a judge-made reasonableness test — duration, geographic reach, burden on the person’s ability to earn a living, whether a real business interest is protected, and the public interest. No safe-harbor number. Some professions, physicians among them, have their own statutory limits. Massachusetts replaced judicial instinct with hard rules in 2018: generally a twelve-month cap, in writing and signed by both sides, express notice of the right to consult counsel, delivery with the offer or ten business days before the start date, and real consideration, often garden leave. Whole categories of worker cannot be bound, including many hourly employees and anyone terminated without cause. Procedural misses are fatal there in a way they are not next door. New York remains common law and a moving target: a broad ban passed the legislature and was vetoed in 2023, and the subject keeps returning. The FTC’s 2024 rule would have banned most non-competes nationwide; a federal court set it aside before it took effect, so it is not in force. Older covenants deserve a fresh read.

A non-solicit is usually the workhorse — narrower, and courts are more comfortable protecting relationships you invested in than keeping someone out of their trade. Confidentiality covenants are the ones that survive when the others do not, provided they define categories rather than claiming “all information” and carry the federal whistleblower notice that pre-2016 templates are missing.

Deliverables: covenants drafted for the enforcing state · a read on whether what you already use would survive review · consideration handled properly, so the covenant is not void on signature · enforcement or defense options when somebody leaves and competes.

Also handled: pursuing a departure that crossed a line, and defending a threatening letter about an agreement that may not hold up · covenants drafted alongside trade secret protection, since they are the same conversation · litigation where it goes that far.

A separation handled properly closes and stays closed. Handled badly it stays open for as long as the statute allows, on the departing employee’s timeline rather than yours.

Severance is generally not required unless a contract, handbook, or policy says so. It matters for a different reason: severance is the consideration that makes a release enforceable. If you want a departing employee to give up claims, you have to give them something they were not already owed.

A release can cover most claims, with limits. Some cannot be waived, and certain categories carry their own consideration and timing requirements a generic template will not satisfy. A release drafted for the wrong state or the wrong age bracket can be worth nothing.

The work also covers what happens before the conversation: documentation and process that limit risk before you let someone go, and a written plan for the meeting itself, including what not to say in it. Reductions in force get structured so selection criteria and notices hold up.

Deliverables: a separation agreement and release drafted for the state that governs it · the consideration and timing requirements that make a release enforceable · confidentiality, non-disparagement, and return-of-property terms · a written plan for the conversation.

Also handled: reductions in force and layoff selection criteria · severance structuring · demand letters and claims that arrive after a separation · recovering company property and data when the exit was not clean.

Business legal audit — one written pass across the whole employment file: offer letters, contractor roster and classification, handbook, covenants, and separation practice, reporting what needs fixing and in what order.

Outside general counsel — an ongoing relationship on a monthly retainer. Hiring and separations rarely keep to a schedule, and this is what the questions in between are for.