How to Decide: Hire or Contract?

Hire or contract is one of the decisions where "the math" gives the wrong answer on its own. The legal classification is the first gate — if a role doesn't pass it, the loaded-cost comparison is moot. This guide is the decision frame: it walks the three gates in order, then hands off to two deep-dives for the parts that reward more rigor (the cost math and the classification and honest tradeoffs).

1. Classification: the IRS and DOL tests

Federal classification tests look at substance over form. Calling someone a contractor doesn't make them one. For employment taxes, the IRS applies the common-law rules, grouping the evidence into behavioral control, financial control and the type of relationship, and says no set number of factors decides it^[1]. The 20-factor list many guides still quote comes from a 1987 IRS ruling (Revenue Ruling 87-41); the three categories are how the IRS organises the same evidence today. For minimum wage and overtime, the Department of Labor applies an economic-reality test under the FLSA.

That federal FLSA test is in flux. The 2024 DOL final rule, effective March 11, 2024, set out a six-factor economic-reality test^[2]. Since May 1, 2025 the Wage and Hour Division no longer applies it in its own investigations, using its older Fact Sheet #13 guidance instead, while the 2024 rule still governs private lawsuits^[5]. On February 26, 2026 the DOL proposed replacing it^[6] with a test built on two core factors, control over the work and the worker's opportunity for profit or loss, similar to its 2021 rule^[7]. The factors below carry weight under every version.

Core factors that weigh toward employee classification:

• Control over work details. You dictate when, where, how, and with what tools the work is done.
• Economic dependence. The worker relies on you for a majority of their income over time.
• Integration into your operations. They sit on your team Slack, attend your meetings, use your project management tools, report to your managers.
• Duration. Ongoing, indefinite relationship rather than a defined project with an end.
• Provision of tools/equipment. You provide the laptop, software licences, and working environment.
• No profit/loss risk. Worker earns a steady rate regardless of project outcomes.

Factors that weigh toward contractor classification:

• Worker has other clients (not economically dependent).
• Worker sets their own hours and methods.
• Defined-scope engagements with clear start and end.
• Worker provides their own tools and equipment.
• Worker bears profit/loss risk — fixed-fee projects, their own overhead, opportunity to earn more through efficiency or less through delays.

If a role is mostly in column A, classifying the worker as a contractor is misclassification. Uncertain cases: file IRS Form SS-8 for a determination^[4], or err toward employee status to avoid back-tax liability.

2. The loaded-cost comparison

For roles that legitimately qualify for either classification, compare loaded costs:

• Employee: Base salary × about 1.3. BLS June 2026 data put employer taxes and benefits at 1.24x to 1.29x of a salary, before equipment, software and office costs^[3] (the derivation is in the true loaded cost of an employee). For a $100k base: ~$130k annual loaded cost, delivering roughly 1,700 productive hours = $76/hour effective rate.
• Contractor: Hourly rate × actual hours. For the same 1,700 hours of work at $100/hour = $170k annual.

At face value, the employee looks cheaper per hour. Three adjustments usually matter:

• Utilisation. Employees are paid whether they are fully utilised or not. If the actual need is 20 hours per week of specialized work, the full-time employee is over-provisioned. A contractor billing only 20 hours at $100 = $104k annual — below the employee loaded cost.
• Ramp time. A new employee takes time to reach full productivity; an experienced contractor hired for one defined skill usually needs less. The shorter the engagement, the larger the share of it that ramp-up consumes.
• Cessation cost. Contractor engagements end cleanly. Employees don't — severance, unemployment insurance experience rating, and the morale cost of involuntary exits all have financial implications.

3. Scope stability is the hidden deciding factor

Employees are efficient per-hour-of-work when the workload is steady. Contractors are efficient when the workload is spiky or uncertain.

The workload question:

• Will this role have enough work for 35–40 hours per week, every week, for the next 12+ months?
• Is the work steady enough that ramp-up time amortises over a long productive period?
• Is the skill set specialised enough that off-hours productivity can't be absorbed by training into adjacent work?

If yes to all three: employee is usually correct. If the work is uncertain, project-based, or the skill set is narrow enough that full-time work would not fill the role productively, contractor is usually correct.

Example. A company needs design work. They could hire a $95k designer full-time or contract a senior designer at $90/hour. If the work is steady and full-time: employee is ~$124k loaded for 1,700 productive hours = $73/hour; a contractor billing the same 1,700 hours at $90/hour costs $153k. Employee wins by ~$30k/year. If the work averages 15 hours/week of specialist output with occasional spikes: contractor is $90/hour × 780 hours = $70k; the employee is still $124k but delivering only modest incremental value in the slack periods. Contractor wins by ~$54k.

4. A decision framework that works

Three questions, in order:

1. Does the role legitimately qualify for contractor classification? If no, the decision is made — it must be an employee or it must be restructured.
2. Is the workload steady at 35+ hours per week for 12+ months? If yes, default to employee. If no, default to contractor.
3. Is there a meaningful strategic reason to deviate? (Specialist scarcity, short runway, building a competitive capability that must be internalized, etc.)

Most roles should land clearly in one category. The ambiguous middle — part-time specialist work, trial periods before full-time conversion, is where most misclassification happens.

5. Common mistakes and misclassification risk

Misclassification is expensive when caught. The IRS can assess the employment taxes that should have been paid, with penalties and interest; an employer with no reasonable basis for treating the worker as a contractor cannot use the relief provisions that would otherwise reduce the bill^[1]. Misclassified workers can also recover unpaid minimum wage and overtime under the FLSA^[2], and state unemployment, wage and tax claims add another layer. The IRS Voluntary Classification Settlement Program offers partial relief to employers that reclassify workers before an audit^[1].

Common mistakes that create exposure:

• Converting a laid-off employee immediately into a contractor doing the same work. High-risk pattern: the paper trail shows the same duties under a new label.
• Treating contractors like employees — requiring them at daily standups, assigning them to your internal tools, managing their hours. This erodes their legitimate contractor status.
• "1099 employees." There is no such category. A person is either a W-2 employee or a 1099 independent contractor; the term "1099 employee" has no legal meaning and usually means the classification is wrong.
• State-level variation. California and New Jersey, among other states, apply ABC tests that presume a worker is an employee unless all three prongs are met, including that the work is outside the usual course of your business^[8][9]. A contractor classification valid federally may fail state law.

In the typical case, the defensible choice is the one that matches the substance of the relationship. If the decision feels like it's bending the rules for cost reasons, it probably is, and the back-tax exposure is rarely worth the cashflow savings.

6. Contractor-to-employee transitions

A common pattern: hire as a contractor initially to assess fit, transition to employee after 3–6 months. Done well, this works. Done poorly, it creates classification risk.

To do it well:

• Set a defined trial period upfront. Written agreement stating the arrangement is project-based with a specific deliverable, not ongoing work.
• Preserve contractor-substance behaviours during the trial. Let them set their hours, use their own equipment, work outside your daily management cadence.
• Pay project fees or defined deliverables, not hourly. Reduces the weight of "economic dependence" and "hourly management" factors in classification tests.
• Convert cleanly at offer time. W-2 onboarding, full benefits, revised compensation that reflects the different loaded-cost economics.

Common mistake: "contract-to-hire" as a euphemism for "employee with delayed benefits." If the person is working 40 hours/week at your office using your tools and reporting to your manager, they are an employee — regardless of what the contract says.

7. International hiring considerations

For remote-first companies, the hire-vs-contract question gets more complex internationally. Key decisions:

• Local entity. If you plan to build a team in one country, your own local entity spreads its setup and ongoing compliance cost across every hire there. Get local quotes for incorporation, payroll, accounting and registered-office costs before comparing it with an EOR.
• Employer of Record (EOR). For a first hire or a few in a country, an EOR (Deel, Remote, Oyster, similar) employs the worker for you and avoids the entity setup. Deel and Remote both list EOR at $599 per employee per month, on top of the salary and the country's employer taxes and benefits (Deel vs Remote EOR pricing).
• International contractors. Paying international contractors is simpler administratively but carries classification risk wherever local labour law looks at substance over form. Some jurisdictions won't recognise a contractor arrangement that looks like employment regardless of what the contract says.

Before hiring internationally, consult local tax and employment counsel. The up-front legal review is usually far cheaper than unwinding a misclassification later. Plan against the strictest test that applies to the worker's location, not the most lenient one.

8. Numeric worked example — 30-hour role across three structures

A bootstrapped SaaS needs a senior designer for ~30 hours/week of product and marketing design. The workload is predictable but doesn't fill a full-time role: it is 75% of the 1,700 productive hours a full-time employee delivers in a year, or about 1,275 hours. Compare three structures delivering those same hours.

Structure          Annual cost     Effective $/hour   Notes
─────────────────────────────────────────────────────────────────────────
FT employee        $115k base × 1.3 = $149.5k loaded
                   / 1,275 hours of work needed =              ~$117/hr
                   Risk: 10 hr/wk slack paid for

Part-time W-2      $86.3k (30/40 pro-rated) × 1.25 = $107.8k loaded
                   / 1,275 hours =                             ~$85/hr
                   Works if role genuinely qualifies at this cadence

Contractor         $95/hr × 1,275 = $121.1k                    $95/hr
                   No benefit load, clean termination
                   Must pass the IRS and DOL tests[1,2]

Part-time W-2 is the cheapest per productive hour and eliminates misclassification risk. FT W-2 at 75% utilisation is the most expensive per actual hour delivered and pays about $37k a year for unused capacity. Contractor sits between the two on pure cost but creates the classification question: 30 regular hours per week on your tools, attending your team meetings, under your day-to-day direction looks like employment under both the IRS and DOL tests^[1][2]. If the work genuinely is defined-scope with worker autonomy, contractor is defensible; if it's "employee with a 1099," it isn't.

9. Failure modes worth naming

• Ex-employee re-engaged as 1099 doing the same work. The highest-risk misclassification pattern, because it leaves a paper record of identical scope under a different classification. If an ex-employee must continue, structure genuinely different scope (defined project, external tools, client-facing autonomy) or accept W-2 continuation.
• State-level ABC tests ignored. California's ABC test is materially stricter than the federal tests^[8], and New Jersey's ABC test regulations took effect on October 1, 2026^[9]. A role that passes IRS scrutiny federally can fail state review. For contractors in ABC-test states, run the state-level test explicitly.
• EOR relationship treated as "same as contractor." Employer-of-Record services (Deel, Remote, Oyster) make the worker an employee of the EOR, not your company, and the fee ($599 per employee per month on Deel's and Remote's list prices) comes on top of the full employment cost. Don't price them against the contractor's rate; price them against the W-2 loaded cost plus the fee, and against the cost of running your own entity in that country.

The federal test is moving: the DOL stopped applying its 2024 rule in its own investigations in May 2025 and proposed a replacement in February 2026^[5][6], while states such as New Jersey have tightened theirs^[9]. A classification that relies on the most lenient current reading can fail when the rule or the jurisdiction changes, so structure the engagement to pass the stricter tests.

Related in this series

This guide is the decision frame. Two companion deep-dives go further on the two parts that need the most rigor:

• FTE vs contractor cost math — the loaded-cost deep-dive: a $90k employee against a $60/hr contractor, built column by column, compared per working hour, with ramp time and the federal tax exposure of misclassification.
• Contractor vs employee, honestly (2026) — the classification-and-tradeoffs deep-dive: IRS reclassification exposure, the flexibility-versus-continuity calls, and where the honest answer diverges from the cheapest one.

Frequently asked questions

Should I hire an employee or a contractor?

Run three gates in order. First, legal classification: does the role legitimately qualify as a contractor under the IRS common-law rules and the Department of Labor's economic-reality test? If not, the decision is made — it must be an employee. Second, loaded-cost comparison: an employee at about 1.3 times base versus a contractor at market hourly rate for the same hours of work. Third, scope stability: will the workload be steady at 35-plus hours a week for 12-plus months? If yes to all, default to employee; if the work is spiky, project-based, or too narrow to fill full-time, default to contractor.

What determines whether a worker is a contractor or employee?

Substance over form — calling someone a contractor does not make them one. Factors weighing toward employee status include your control over when, where, and how the work is done; the worker's economic dependence on you for most of their income; integration into your operations (your Slack, meetings, tools, managers); an ongoing indefinite relationship; you providing the equipment; and no profit-or-loss risk. Contractor factors are the reverse: multiple clients, self-set hours and methods, defined-scope engagements, own tools, and genuine profit-or-loss risk. For uncertain cases, file IRS Form SS-8 or err toward employee status to avoid back-tax liability.

When is a contractor cheaper than an employee?

When utilization is low or the engagement is short. An employee is paid whether fully utilized or not, so if the real need is 20 hours a week of specialized work, a contractor billing only those hours at $100 (about $104k a year) comes in below an employee's loaded cost of about $130k on a $100k salary. An experienced contractor also usually needs less ramp-up than a new hire, which matters most on short engagements, and contractor engagements end cleanly while employees carry severance and unemployment-rating costs. For steady 35-plus-hour-a-week work, the employee is usually cheaper per productive hour.

What is the risk of misclassifying a contractor?

It is expensive when caught. The IRS can assess the employment taxes that should have been withheld and paid, with penalties and interest, and an employer with no reasonable basis for the classification loses the relief provisions. The Department of Labor and private plaintiffs can recover unpaid minimum wage and overtime, and states add their own unemployment, wage and tax claims. The highest-risk pattern is converting a laid-off employee straight into a 1099 doing the same work. Note there is no such thing as a 1099 employee, and state ABC tests, such as California's and New Jersey's, can fail a classification that passes federally, so run the state-level test explicitly in those jurisdictions.