Leading vs laggingsafety indicators in construction.
Lagging indicators count incidents after they happen. Leading indicators predict them. The 2026 reference for construction safety KPIs, with ANSI Z10, OSHA 3885, and CPWR-validated metrics.
Lagging safety indicators measure outcomes after incidents have already happened like TRIR, DART, LTIR, EMR, fatalities, and OSHA citations. They are reactive, required by OSHA recordkeeping, and central to prequalification scoring. Leading indicators measure activities before incidents happen. Training completion, near-miss reporting, hazard observations, JSA completion, audit closure. They are proactive and predictive of future recordable rates. A balanced construction safety scorecard combines both, with 60 to 70 percent of management attention on leading indicators (controllable today) and 30 to 40 percent on lagging indicators (required for compliance and benchmarks).
The 10 most common construction leading indicators
Each measures an activity that happens before incidents. Pair every volume metric with a quality companion to prevent gaming.
Training completion rate
- Formula
- Trainings Completed / Trainings Scheduled
Percentage of required safety trainings completed against the planned roster. Tracks whether workers reach the jobsite with the certifications and refreshers they need.
Near-miss reporting rate
- Formula
- (Near-Miss Reports x 200,000) / Hours Worked
Number of near-miss reports per 200,000 hours worked. Counterintuitively, a higher rate signals stronger safety culture, not worse performance. Best-in-class contractors target 20 to 40.
Hazard observation rate
- Formula
- Observations Submitted / Crews / Week
Documented hazard observations per crew per week. Captures the everyday spotting of unsafe conditions before they cause harm. Pair with action closure rate to prevent paperwork inflation.
JSA / pre-task plan completion rate
- Formula
- JSAs Completed / Tasks Started
Percentage of tasks that had a documented Job Safety Analysis or pre-task plan completed before work began. CPWR research flags this as one of the strongest predictors of recordable reduction.
Toolbox talk attendance and quality
- Formula
- Attendance Rate x Quality Score
Toolbox talk attendance multiplied by a quality score (topic relevance, worker engagement). Counting talks held without quality scoring incentivizes empty meetings.
Audit and inspection completion rate
- Formula
- Audits Completed / Audits Scheduled
Percentage of safety audits and inspections completed against the scheduled cadence. Pair with action closure rate so audits drive change, not just paperwork.
Stop Work Authority (SWA) exercises
- Formula
- SWA Events / Period
Count of times any worker exercised stop-work authority. A healthy SWA exercise rate is a leading indicator of a mature culture where workers trust they will be supported, not punished.
PPE compliance audit rate
- Formula
- Compliant Observations / Total Observations
Percentage of randomly sampled PPE observations that meet site requirements. Best as a sampled audit, not a binary yes/no for every worker, every day.
Behavior-Based Safety (BBS) observation rate
- Formula
- BBS Observations / Worker / Month
BBS observations submitted per worker per month. Tracks the volume of peer-to-peer safety conversations happening on site, which research links to reduced at-risk behavior over time.
Action Required Rate
- Formula
- Deficiencies Identified / Audit
Average count of deficiencies identified per safety audit. A leading indicator that surfaces emerging hazards earlier. Pair with average days-to-close to measure responsiveness.
The 6 required construction lagging indicators
These appear on OSHA 300A summaries, owner prequalification packets, and workers comp policies. The lagging side of every construction safety scorecard.
TRIR (Total Recordable Incident Rate)
- Formula
- (Recordable Cases x 200,000) / Hours Worked
OSHA-recordable injuries and illnesses per 200,000 hours worked. Required for the 300A summary. The single most-quoted construction safety metric.
DART (Days Away, Restricted, Transferred)
- Formula
- (DART Cases x 200,000) / Hours Worked
Recordable cases that involved lost time, restricted duty, or job transfer. A subset of TRIR. Heavily weighted in owner prequalification scoring.
LTIR (Lost Time Injury Rate)
- Formula
- (Lost Time Cases x 200,000) / Hours Worked
Injuries that caused at least one full day away from work, per 200,000 hours worked. Used internationally and on owner safety dashboards.
EMR (Experience Modification Rate)
- Formula
- Actual Losses / Expected Losses (3-year window)
Workers compensation premium multiplier set by NCCI or an independent state bureau. Three-year window. The most heavily weighted prequalification metric.
Fatality count
- Formula
- Count per period
Work-related fatalities in the measurement period. The most consequential lagging indicator. Always reported separately, never blended into rates.
OSHA citations and penalties
- Formula
- Citation count, penalty dollars
Number of OSHA citations issued and dollar value of penalties in the measurement period. Tracked separately for serious, willful, and repeat citations.
Leading vs lagging at a glance
The two indicator families answer different questions, run on different cadences, and live in different places on the management calendar.
| Attribute | Leading | Lagging |
|---|---|---|
| What it measures | Activities before incidents | Outcomes after incidents |
| Posture | Proactive, predictive | Reactive, descriptive |
| Time horizon | Today and this week | Last month, last quarter, last year |
| Controllability | High, managers act on these directly | Low, already in the past |
| Required by OSHA | No (recommended by 3885) | Yes (29 CFR 1904) |
| Used for prequalification | Increasingly, for advanced programs | Yes, standard on every packet |
| Where management attention belongs | 60 to 70 percent of focus | 30 to 40 percent of focus |
| Update cadence | Weekly at site level | Monthly company, annual EMR |
How to build a balanced construction safety scorecard
Five steps. ANSI Z10 and OSHA Recommended Practices 3885 as the framework. CPWR research for indicator selection.
- 1
Select your lagging baseline
Start with TRIR, DART, LTIR, and EMR.
These four lagging indicators are required for OSHA recordkeeping and appear on essentially every owner prequalification packet. Pull twelve months of historical data so you have a baseline to measure improvement against. Compare to BLS construction benchmarks (NAICS code 23) and to any owner-specific thresholds in your prequalification documents.
- 2
Choose three to five leading indicators
Pick from the CPWR-validated list.
The strongest predictors of recordable reduction in construction research are pre-task plan completion rate, near-miss reporting rate, and supervisor safety observation rate. Training completion rate and audit closure rate are also strong. Three to five is the sweet spot. More dilutes management attention. Less misses leading-indicator coverage of major risk areas.
- 3
Pair every volume metric with a quality metric
Prevent gaming by pairing count with closure.
Count of hazard observations without an action-closure rate incentivizes paperwork that nobody resolves. Count of toolbox talks without quality scoring incentivizes empty meetings. For every volume leading indicator, define a quality companion: action closure rate, days-to-close, recurrence rate, or quality audit score.
- 4
Set targets and review cadence
Weekly site, monthly company, quarterly board.
Define quarterly targets for each indicator based on twelve-month history and industry benchmarks. Review leading indicators weekly at the site or project level, where supervisors can act on them. Review lagging indicators monthly at the company level, where trends become statistically meaningful. Present the full scorecard quarterly to executives and annually for ANSI Z10 management review.
- 5
Validate with cross-checks
Confirm leading trends correlate with lagging outcomes.
Every quarter, verify that your leading indicators are actually moving the lagging outcomes. If near-miss reporting is rising but TRIR is also rising, investigate whether observations are surfacing problems faster than the team can resolve them. If both move in the right direction, the scorecard is working. If leading indicators are flat but lagging deteriorates, the scorecard is the wrong shape and needs redesign.
Worked example: a balanced scorecard in action
A mid-size general contractor with 240 employees and 480,000 hours worked annually introduces a leading-indicator program. Twelve months later, the lagging numbers follow.
The starting point looks average. Lagging indicators sit slightly worse than the BLS construction baseline. Leading indicators barely exist, because the company was only tracking what OSHA required.
- TRIR
- 3.2
- DART
- 1.4
- EMR
- 0.95
- Near-miss reports
- 12 / yr
- Hazard observations
- Not tracked
- Training compliance
- Unknown
- TRIR
- 1.8
- DART
- 0.7
- EMR
- 0.87
- Near-miss reports
- 184 / yr
- Hazard observations
- 8 / crew / wk
- Training compliance
- 96%
Near-miss reports jumped from 12 to 184. That looks alarming in isolation, but it is the leading-indicator signature of a stronger reporting culture, not worse safety. Hazard observations and training compliance climbed alongside it. Twelve months later, every lagging indicator has improved: TRIR dropped from 3.2 to 1.8, DART from 1.4 to 0.7, EMR from 0.95 to 0.87. More leading data, fewer lagging incidents. That is the pattern ANSI Z10 and CPWR research describe.
Five mistakes that wreck a safety scorecard
These appear in every safety program audit. Each one undercuts the predictive value of the leading-indicator side and obscures what is actually happening in the field.
- 01
Tracking only lagging indicators
Outcome-only management means waiting for someone to get hurt before you have data to act on. By the time TRIR has moved a tenth of a point, the incident that caused it happened weeks or months ago. Leading indicators let you intervene before harm. A scorecard with only lagging metrics is a rear-view mirror.
- 02
Punishing high near-miss reporting
A rising near-miss reporting rate is good news, not bad. Sites with zero near-miss reports are almost always sites with suppressed reporting, not safe sites. If supervisors get penalized for high near-miss counts, the workforce learns to stop reporting. The data goes dark, and the next recordable is a surprise.
- 03
Counting observations without closing them
Submitting a hazard observation that never gets resolved is theater. Best-in-class programs pair observation rate with an action closure rate target (typically 80 to 95 percent closed within a defined window) and an average days-to-close metric. Volume without closure incentivizes paperwork that nobody acts on.
- 04
Letting toolbox talks become check-the-box
A toolbox talk attendance rate of 100 percent is meaningless if the talks are five minutes of reading the script. Pair attendance with a quality dimension: topic relevance to current work, worker engagement, action items generated, and follow-through on those action items.
- 05
Comparing across unrelated industries
A 2.0 TRIR is well below average for construction (NAICS 23, BLS reports 2.5) and well above average for office work. Comparing your construction safety scorecard to manufacturing or general industry numbers is meaningless. Compare to NAICS 23 or to your specific construction sub-sector.
Industry benchmarks for construction
Compare against NAICS code 23 (construction), not against manufacturing or general industry. Numbers below are sourced from BLS, CPWR, OSHA, and CII publications.
Frequently asked questions
Building a balanced safety scorecard is straightforward in theory and brutal in practice. Near-miss reports live in one binder. Toolbox talks live in another. Hazard observations are written on the back of the daily report. JSAs sit in an email thread. By the time anyone assembles the leading indicators, the month is over and the lagging numbers are the only thing left to react to. Plan of Day is voice-first construction reporting that captures near-misses, hazard observations, JSAs, training records, and audit findings as field reports land, routes them through POD's intelligence engine, and feeds hundreds of KPIs that include the full leading and lagging indicator suite. Specialized AI agents flag at-risk trends before they become recordables. The scorecard updates continuously, not at month end.
Further reading
TRIR vs EMR vs LTIR vs DART
The four lagging safety rates explained: formulas, benchmarks, calculation mistakes, and how each is used in prequalification.
OSHA 300 Log Requirements for Construction
Recordkeeping rules under 29 CFR 1904. What counts, who must keep it, and how the 300 log feeds your lagging indicators.
Construction Daily Report Guide
How safety observations on the daily report become leading indicators: near-miss reporting, hazard observation, JSA completion.
POD for Safety Managers
How safety managers use POD to track leading and lagging indicators continuously instead of assembling them by hand.
Sources
- ANSI/AIHA Z10-2019 . U.S. consensus standard for Occupational Health and Safety Management Systems, including leading and lagging indicator requirements.
- OSHA Recommended Practices for Safety and Health Programs in Construction (3885) . Voluntary seven-element framework, current edition 2022.
- U.S. Bureau of Labor Statistics, Survey of Occupational Injuries and Illnesses (SOII) . Annual incident-rate data by NAICS code for construction benchmarks.
- Center for Construction Research and Training (CPWR) . Twenty-plus years of construction leading-indicator research and validated predictor lists.
- Construction Industry Institute (CII) . Best Practices for Safety, scorecard design guidance, leading-to-lagging ratio recommendations.
- OSHA Voluntary Protection Programs (VPP) . Qualification criteria emphasizing leading-indicator measurement.