Series 2 · Employment classification

The Installers on Form 1099

Q34Four modelsSix blind pairingsMaterial-outcome scoring

Why this result mattered

Functional tie

The principal comparison placed Qwen 3.8 and Claude under concealed A/B labels.

Both answers reject a zero reserve, calculate the same $354,922.40 nonwillful reserve, identify element B as likely fatal, and direct immediate preservation and factual investigation before conceding classification. Both lead Northbank toward prospective correction and a fact-supported closing agreement or limited contest, with no demonstrated difference in the company’s material position.

Benchmark materials

Prompt and sealed key

The key is an audit reference. It was not shown to the contestant systems and did not determine the comparative verdict.

Read the original matter prompt

Question 34: The Installers on Form 1099

Your role

You represent Northbank Solar, which treated field installers as independent contractors for two tax years in the fictional State of Daxton. The revenue agency is asking questions. Management needs a conditional classification analysis, a computed reserve, and a plan to obtain control facts not yet gathered.

Governing law

Assume the following Daxton rules apply:

  1. A worker performing services for pay is presumed to be an employee unless the business proves all three elements: (A) freedom from control in contract and practice; (B) work outside the usual course of the business; and (C) an independently established trade of the same nature.

  2. Written contractor labels, invoices, and tax forms are relevant but do not control. Actual scheduling, supervision, pricing, tools, substitution rights, profit opportunity, and work for others are relevant to elements A and C.

  3. Installation of the product a company sells is within the company's usual course when installation is marketed as part of the customer offering and produces a substantial share of revenue.

  4. If reclassified, the employer owes an 8 percent employer payroll tax and a 7 percent employee withholding tax on covered compensation.

  5. When the employer timely filed contractor information forms, it receives a credit equal to 40 percent of the otherwise unwithheld 7 percent employee tax. The employer may not seek retroactive reimbursement from workers.

  6. A nonwillful classification failure carries a penalty equal to 10 percent of the net unpaid tax after the rule 5 credit. A willful failure carries 30 percent. The supplied record establishes the nonwillful rate for reserve purposes unless later facts show concealment.

  7. For a reserve calculated as of December 31, 2026, interest is fixed at 4 percent of net 2025 tax and 2 percent of net 2026 tax. Do not compound.

  8. A prospective change does not admit past liability. Backdating contracts, scripts, or time records is prohibited.

  9. The agency may approve a closing agreement after accurate disclosure of worker counts, compensation, and practices. Counsel should not concede employee status before gathering the facts needed for element A.

Facts

Northbank sells residential solar systems and advertises a single price for equipment and installation. Installation generated 38 percent of gross revenue.

Northbank used:

  • Fourteen installers in 2025, paid total covered compensation of $980,000.
  • Twenty installers in 2026, paid total covered compensation of $1,600,000.

Northbank timely issued contractor information forms for every payment. No payroll tax or withholding was paid on these amounts.

The written agreements call the installers independent contractors, require their own hand tools and insurance, and permit work for other companies. Most installers formed small businesses and filed business tax returns. Northbank provides the solar panels, inverters, safety plans, and customer contracts.

Management has not yet gathered reliable facts about who set daily arrival windows, whether installers could reject jobs without penalty, whether they could send qualified substitutes, who controlled work sequence, whether Northbank supervisors directed methods onsite, how often installers worked for competitors, or whether installers could negotiate price. Emails and dispatch records exist but have not been reviewed.

The chief operating officer says, "They signed contractor agreements, so reserve zero." The controller asks for the maximum nonwillful reserve if all compensation is reclassified, using only the supplied rates. The agency has not issued a formal assessment.

Assignment

Analyze:

  • The likely treatment of element B and the facts relevant to elements A and C.
  • Why the written labels do or do not resolve status.
  • The exact 2025 employer tax, net employee tax after credit, penalty, and interest.
  • The exact 2026 employer tax, net employee tax after credit, penalty, and interest.
  • The combined nonwillful reserve if all compensation is reclassified.
  • The specific control and independent-trade facts that must be obtained before a final classification position.
  • Immediate document-preservation and interview steps.
  • Whether to change practices prospectively, seek a closing agreement, contest classification, or defer the decision pending the missing facts.


Read the sealed key

Q34 key (Installers on Form 1099)

  1. Element B strongly favors employee status because installation is sold as part of Northbank's offering and produces 38 percent of revenue. The agreements do not overcome that statutory fact. Missed: the company treats labels as dispositive.
  2. Elements A and C cannot be finalized without scheduling, rejection, substitution, supervision, methods, outside-work, and price-negotiation facts. The small businesses, tools, insurance, and contractual permission help Northbank but do not answer actual practice. Missed: counsel concedes or denies status before gathering the controlling evidence.
  3. For 2025: employer tax is 8 percent of $980,000 = $78,400. Gross employee tax is 7 percent = $68,600; the 40 percent credit is $27,440; net employee tax is $41,160. Net tax is $119,560; 10 percent penalty is $11,956; 4 percent interest is $4,782.40. Total 2025 reserve is $136,298.40. Missed: the reserve is materially wrong.
  4. For 2026: employer tax is 8 percent of $1,600,000 = $128,000. Gross employee tax is $112,000; credit is $44,800; net employee tax is $67,200. Net tax is $195,200; penalty is $19,520; 2 percent interest is $3,904. Total 2026 reserve is $218,624. Missed: same.
  5. Combined nonwillful reserve is $136,298.40 + $218,624 = $354,922.40. Missed: management carries zero or an unsupported rounded exposure.
  6. Preserve dispatch, pricing, supervisor, substitution, customer, invoice, and outside-work records; interview a representative cross-section before agency admissions. Missed: the actual-practice evidence is lost or filtered through executives.
  7. A prospective compliant redesign may proceed without admitting the past, but no backdating. Reserve the computed maximum, gather facts, then decide whether to contest elements A and C or pursue a closing agreement. Missed: the company either freezes a flawed model or concedes before it knows its best defense.
Blind evaluation record

Direct judge report

The report below is preserved from the Qwen 3.8 versus Claude verdict record.

Read the preserved judge report
TIE, FUNCTIONALLY EQUIVALENT.

Basis. Both answers reject a zero reserve, calculate the same $354,922.40 nonwillful reserve, identify element B as likely fatal, and direct immediate preservation and factual investigation before conceding classification. Both lead Northbank toward prospective correction and a fact-supported closing agreement or limited contest, with no demonstrated difference in the company’s material position.

[A] rejected: loss-of-credit sensitivity — record establishes timely contractor information forms.
[A] rejected: document alteration automatically converts the penalty from 10% to 30% — rule requires later facts showing willfulness or concealment.
[B] rejected: alternative interest calculation including penalties — governing rule fixes interest on net tax, not penalties.

Failed candidate: timing of prospective changes — different conduct failed because Answer A’s implementation timeline follows investigation, materially matching Answer B.
Failed candidate: reserve-booking language — record support failed because no governing accounting standard was supplied.
Failed candidate: scope of any classification contest — concrete consequence failed because both limit a contest to factually supportable workers or services.
Failed candidate: additional fact requests and interview instructions — concrete consequence failed because no different classification, disclosure, or reserve outcome is demonstrated.

Hard stop. The required sentence cannot be completed with a concrete, record-supported consequence.
Complete comparison record

All six pairings

Every pair of the four model answers was judged independently under the same evaluation instrument.

Open the six head-to-head results
PairingResultWhy it matteredRecord
Qwen 3.8 27B vs Claude Opus 5Functional tieBoth answers reject a zero reserve, calculate the same $354,922.40 nonwillful reserve, identify element B as likely fatal, and direct immediate preservation and factual investigation before conceding classification.Judge record
Qwen 3.8 27B vs DeepSeek V4-Flash-0731Functional tieBoth answers direct Northbank to reject the zero-reserve position, record the same $354,922.40 conditional nonwillful reserve, preserve and investigate the missing A-and-C facts, avoid conceding status prematurely, and then consider prospective reclassification or a closing agreement.Judge record
Qwen 3.8 27B vs Qwen 3.5 27BFunctional tieBoth answers reject the zero-reserve position, conclude element B likely fails, calculate the same $354,922.40 nonwillful reserve, and direct immediate preservation and fact-finding on elements A and C before conceding status or choosing between contest and a closing agreement.Judge record
DeepSeek V4-Flash-0731 vs Claude Opus 5Functional tieBoth answers direct Northbank to reserve $354,922.40 now, preserve evidence, investigate the missing A and C facts, avoid conceding status prematurely, consider prospective correction, and evaluate a closing agreement after accurate fact development.Judge record
DeepSeek V4-Flash-0731 vs Qwen 3.5 27BFunctional tieBoth answers reject a zero reserve, calculate the same $354,922.40 nonwillful reserve, treat element B as likely failed, and direct Northbank to preserve records and investigate elements A and C before conceding status or choosing a final agency position.Judge record
Qwen 3.5 27B vs Claude Opus 5Functional tieBoth answers conclude that element B likely fails, reject reliance on contractual labels, calculate the same $354,922.40 reserve, and direct preservation and investigation of control and independent-trade facts before any concession.Judge record
Rejected-point diagnostic

Judge-attributed rejected points for this matter across all six pairings:

Qwen 3.8 27B4
DeepSeek V4-Flash-07310
Claude Opus 58
Qwen 3.5 27B3

This is a review-burden diagnostic, not a universal hallucination measure.

Available reasoning traces