How General Motors Pays

Salaried professional ladder — Level 5 through Level 9

Leadership ladder — Director through Senior Vice President

Disclosed executive layer

Track divergence

Hierarchy qualifications and legacy structures

• GM postings and recurring employee reports identify Levels 5 through 9, sometimes with A, B and C subgrades, but no public source establishes a complete one-to-one map for every function, country or legacy operation.
• This is not a software-only ladder. Manufacturing group leaders, engineers, software developers, managers, technical specialists, chief engineers and fellows can occupy the same numbered levels, so title alone is an unreliable compensation proxy.
• Levels 7 to 9 are the most ambiguous rungs. Role family, subgrade, geography and business criticality can matter as much as the headline number, which is why the peer mapping here is a functional comparison and not a grade equivalence.
• Director, Vice President, Senior Vice President, Executive Vice President and Chief Executive Officer are corporate titles, not hidden grade numbers, and are treated as title tiers throughout this report.
• The unionised hourly population sits outside this ladder entirely. It is covered by negotiated classifications and progression rather than salaried levels, and the source bundle carries no collective-agreement wage table, so no hourly scale is shown.
• Country entities and joint-venture environments, particularly GM Korea and GM China, can use local titles that do not map cleanly onto the US Level 5 to Level 9 labels.

Peer-level mapping

Office and market catalogue — calibration factors versus Warren

GM does not publish city-level headcount, so this catalogue does not rank sites by employee count. The Warren Technical Center is the one site GM quantifies, at more than 21,000 employees; GM de Mexico reports more than 25,000 direct employees across five locations and GM Canada approximately 6,900. Oakville was dropped from the location list because the source bundle records no GM headcount there.

Warren anchor medians — the basis of every modeled cell

• The Warren column is the anchor. Its Level 5 median is calibrated on the Flint Level 5 posting, its Level 7 and Level 8 medians on live GM software and engineering-manager postings, and its Level 9 and Director medians on employee-contributed management records.
• Detroit sits 3 percent above Warren, Mountain View 55 percent above and Austin 15 percent above, while Fort Wayne at 0.88 is the lowest-cost US market modelled. The US spread from Fort Wayne to Mountain View is 1.76 times.
• Offshore compression is strongly level-dependent in India, Mexico, Brazil, China and Argentina, so those markets carry separate junior and senior factors. Bangalore runs about 0.09 times Warren at Level 5 and 0.36 times at Senior Vice President, so the gap narrows from roughly eleven times to under three.
• Canada, South Korea and Australia are given a single flat factor because their ratio to Warren moves less than about 15 percent across the whole ladder and is not monotonic. Canada is in fact relatively stronger at junior levels than senior ones.
• Executive Vice President, Chief Executive Officer and Board figures are single global US proxy disclosures that the source bundle repeated unchanged in every city table. They are carried in the executive section instead of the band table so that no city factor is ever applied to them.

Executive incentive targets — percent of salary

Named-executive outcomes

• Stock compensation expense was approximately $334 million in 2025, $332 million in 2024 and $340 million in 2023, with roughly $267 million of unrecognised cost at the 2025 year end. On $185.0 billion of 2025 revenue that expense is about 0.18 percent, which is the clearest single indicator that equity is not a broad-based element of GM pay.
• No stock options were issued in 2024 or 2025, although the plan retains option and stock appreciation right authority.
• Proxy overhang was 8.4 percent and the three-year average burn rate approximately 0.9 percent on the company's own calculation.
• The two utilisation figures below are analytical ratios computed from the disclosed outstanding and available share counts. They are not a company-defined metric.

2020 Long-Term Incentive Plan reserve

Vesting — common reported employee schedule

This schedule describes disclosed named-executive award terms. GM publishes no vesting schedule for the selective awards modelled at Level 7 and above, and the plan and award agreements also carry service, career-status, clawback and restrictive-covenant provisions whose outcome depends on age, service and termination circumstances.

Eligibility by hierarchy level

• No recurring grant was identified at Level 5 or Level 6 in any public source. Spot and discretionary recognition awards may exist, but they are not a band entitlement.
• Level 7 and Level 8 grants are modelled as selective critical-skill, retention or promotion awards, with a modelled median of $10,000 at Level 7 and $40,000 at Level 8 against a low case of zero. They are more common for scarce software, autonomy and senior-management talent.
• Level 9 is the first level modelled with a non-zero low case, at $25,000, rising to a $90,000 median and $200,000 high.
• Director through Senior Vice President grants are modelled as annual Long-Term Incentive Plan awards with medians of $250,000, $900,000 and $3,000,000, on discretionary decisions about performance, role scope, succession and retention.
• Named-executive award mix is disclosed and is generally 75 percent performance share units and 25 percent restricted stock units.
• At the reference share price of $85.80 on 26 August 2026, the modelled medians correspond to roughly 117 shares at Level 7, 466 at Level 8, 1,049 at Level 9, 2,914 at Director and 34,965 at Senior Vice President. These are illustrative share equivalents, not grant counts.

Indicative annual grant value by band — Warren

Annualized planning value (±25%), not the face value of every new-hire grant. Share equivalents use the $85.8 reference price at 26 August 2026 and ignore plan valuation rules and PSU performance. Highlighted rows are disclosed grant-date stock awards.

Named executive target equity

Two of the five 2025 rows are partly reconstructed. The source bundle reproduced full component tables for Mary Barra, Mark Reuss and Paul Jacobson but only salary, cash incentive and total for Sterling Anderson and Rory Harvey, so the stock figure shown for those two is the residual after salary and cash incentive and their all-other-compensation is shown as zero rather than guessed. Sterling Anderson's 2025 total also includes a substantial new-hire element reported at up to $24m of additional value through 2027 and is not representative of a normal Executive Vice President year.

Board compensation framework

The board package is a governance retainer, not an employment band, and the employee chair receives no separate director pay. The deferred share unit element means roughly $204,000 of a standard director's annual retainer moves with the GM share price rather than being paid in cash.

Not publicly disclosed

Modeled promotion planning timeline

The five-tier system reported in August 2024 places roughly the top 5 percent of the cohort in Significantly Exceeds Expectations with a reported 150 percent bonus treatment, about 70 percent in Achieves Expectations at 100 percent, and roughly the bottom 5 percent in Does Not Meet Expectations at zero and subject to stronger performance action. These are reported design intentions, not a permanent contractual forced distribution, and the two intermediate tiers have no published share or factor.

Dataset summary

City-level H-1B wage history

All-years aggregates. P25/P75 are rounded reconstruction values marked modeled; medians and record counts are the stronger fields.

Selected title / worksite records

Evidence classification

Explicit “Not publicly disclosed” index

Known gaps and diligence before relying on a cell

1. 1Only one research bundle exists for General Motors, so no figure on this page could be cross-checked against a second independent compilation. Where the bundle itself flagged an internal inconsistency, such as GM's rounded 88,000 hourly and 68,000 salaried figures not summing to 155,000, that is noted rather than reconciled.
2. 2The Executive Vice President, Chief Executive Officer and Board rows are deliberately excluded from the band table. In the source bundle those three rows were single US proxy disclosures repeated unchanged in all 32 city tables, so including them would have forced every location's factor towards the top of the ladder and produced absurd offshore cells such as an Indian executive priced off a US chief executive salary. They are carried in the executive section instead, and the band table stops at Senior Vice President.
3. 3The Intern and Co-op row was also dropped from the band table. Its offshore factors behave nothing like the salaried ladder, at roughly 0.04 times Warren in Bangalore against 0.09 at Level 5, and folding it in degraded the fit of every other market by about a third. Reported intern rates near $48.75 an hour in Warren, $50 in Mountain View and $61.15 for some Sunnyvale master's placements sit outside the model.
4. 4Total factors are set equal to base factors throughout. In the source bundle base and total factors were identical at Level 5 in all 31 markets and diverged only higher up, because a flat US-dollar equity value had been applied to every US city; leaving that in would have implied that a Mountain View Senior Vice President is paid 1.15 times Warren when their base is 1.55 times it.
5. 5Band factors are interpolated linearly by seniority between the junior and senior anchors. Against the bundle's own per-band figures this understates mid-ladder compression by up to roughly 30 percent at Level 9 in India and about 23 percent at Level 7 in Mexico. The Level 7 ratio is the most reliable single cross-market comparison and matches the bundle's published arbitrage table.
6. 6No employer benefit load is capitalised into total compensation. Base plus target bonus plus modelled annual equity reconciles exactly to the total on every band row.
7. 7Sterling Anderson's and Rory Harvey's stock award figures are residuals after salary and cash incentive, because the source bundle reproduced only partial component tables for them. Their all-other-compensation is shown as zero rather than estimated, so their stock figure is overstated by whatever that amount actually is.
8. 8Oakville was dropped from the location list. The source bundle recorded it as a regional labour-market benchmark with no GM headcount evidence, leaving 31 markets rather than 32.
9. 9Toyota is omitted from the peer chief executive comparison because Kenta Kon took the role on 1 April 2026 and no comparable full-year total existed at the report date. The Stellantis figure is partial-year and is labelled as not comparable.
10. 10The Canadian Unifor tentative agreements of 22 August 2026 are the largest known unknown on this page. They cover more than 4,600 Ontario workers and their economic terms were withheld pending ratification meetings on 29 and 30 August, after the report date, so no Canadian hourly wage outcome is reflected anywhere here.
11. 11H-1B companywide median and quartiles are not published, and the Warren and Milford quartiles are compact summaries of cited example filings rather than complete city distributions. Record counts shown as zero mean the aggregator displayed no count.

FX rates used — 1 USD equals, snapshot 2026-08-26

Single-date conversion layer for comparability; it ignores payroll-date FX, tax, purchasing power, benefits valuation and hedging. Compensation intelligence, not legal, tax, investment, immigration or employment advice.

Source register — 37 sources

The fall from about 162,000 to 155,000 is roughly 4.3 percent in a single year, and it followed two rounds of salaried reductions in 2024 rather than a manufacturing contraction. GM publishes no comparable global voluntary-attrition rate, so the headcount series is the only workforce trend that can be tracked from public filings.