Program Monitoring, Safety Targets, and Plan Updates

Program Monitoring, Safety Targets, and Plan Updates

Monitor the chain from resources to outcomes

A safety program needs evidence about both delivery and results. Inputs include funding, staff, and equipment. Outputs include installed treatments, inspections, training, or enforcement activities. Intermediate outcomes include speeds, yielding, restraint use, or functioning safety devices. Final outcomes include fatalities and serious injuries. Each answers a different question.

Installing ten crossings demonstrates delivery, not a measured injury reduction. A lower observed speed can support the intended mechanism without establishing the program's full effect on fatalities. A falling fatality count can reflect program effects, exposure changes, background trends, or random variation. Use several appropriate measures and a credible evaluation method rather than claiming that one favorable indicator proves the whole program succeeded.

Assign a measure's definition, data source, responsible owner, update frequency, baseline, target or milestone, and reporting format. Use stable definitions and explain data lags. Leading indicators can identify problems early: a rising proportion of nonfunctioning beacons calls for maintenance before sufficient injury data exist for a long-term evaluation.

Know the five federal PM1 safety measures

Under current 23 CFR 490 Subpart B, state DOTs establish annual targets for five safety measures, expressed as five-year rolling averages:

Fatality data use FARS; serious-injury data come from state reporting with the applicable definition. Vehicle-miles data support the rate measures. The nonmotorized measure combines its defined fatal and serious-injury counts; it is not a pedestrian-only rate. Statewide PM1 measures address public roads rather than only state-owned highways.

A rolling average smooths some annual variability but responds slowly to recent changes. Always identify the years represented. A five-year value reported this year is not the current year's preliminary count, and overlapping rolling averages share four years of data. Treating successive values as independent observations can mislead trend interpretation.

Calculate rolling averages correctly

For count measures, sum the five annual counts and divide by five. For rate measures, first calculate each year's rate using its own fatalities or serious injuries and VMT, then average the five annual rates. Do not replace that procedure with the ratio of five-year total fatalities to five-year total VMT; that alternative is a VMT-weighted rate and can differ from the required average.

Suppose annual fatalities are 10, 12, 8, 14, and 11, while VMT is respectively 1.0, 1.2, 1.0, 1.4, and 1.1 billion miles. Annual fatality rates per 100 million VMT are 1.0, 1.0, 0.8, 1.0, and 1.0. Their five-year average is 0.96. The average fatality count is 11. Combining totals instead gives $55/57=0.9649$ fatalities per 100 million VMT, which is close here but is a different calculation.

Check units carefully: one billion miles equals ten units of 100 million miles. A rate of 1.0 per 100 million VMT is not 1.0 per million VMT. Retain adequate precision for calculations and use the applicable reporting precision at the end.

Distinguish state and MPO target responsibilities

States coordinate target setting with related highway safety planning and report annual targets through the HSIP reporting process. Metropolitan planning organizations establish targets within 180 days after the state establishes them. For each measure, an MPO may agree to support the state target or establish a specific numerical target for its planning area, following the applicable requirements.

This does not require every MPO to predict an annual fatality count independently when it elects to support the state target. It still requires integration of safety goals and performance into planning. Projects, plans, and funding choices should explain how they contribute to the adopted targets.

Federal target procedures differ from a local project's evaluation and from a statewide SHSP update. A project study might examine pedestrian injury crashes at selected crossings, while PM1 tracks statewide rolling measures. A statewide target result does not isolate the causal effect of that particular crossing program.

Understand significant progress

FHWA determines that a state has met or made significant progress when at least four of the five measures meet the applicable test. For each measure, the actual performance must be at or below the target, or below the baseline. The alternative conditions apply measure by measure; this is not a rule that meeting one baseline condition excuses the other four measures.

For example, three measures meeting their targets plus one different measure below its baseline can satisfy four qualifying measures, even if the fifth fails. Three qualifying measures are insufficient. Equality qualifies for the target comparison, but the baseline alternative requires performance below, rather than equal to, baseline.

If a state does not meet or make significant progress, current rules require specified use of obligation authority only for HSIP projects and submission of an implementation plan. The relevant amount is tied to the HSIP apportionment for the fiscal year before the target year. Consult FHWA's performance-measure questions and answers for the determination and compliance process. This is not a general cancellation of all transportation funds.

Use monitoring to improve programs and plans

Review delivery barriers and outcomes with responsible partners. If outputs are delayed, address staffing, procurement, agreements, or maintenance. If outputs were delivered but intermediate behavior did not change, examine the intervention and audience. If injuries remain inconclusive, assess whether enough information exists and whether background changes have been addressed.

Record lessons and revise actions, resource allocations, and evaluation methods. Formal plan updates should incorporate new evidence, changed risks, and implementation experience, while routine adjustments can occur between updates. Publish progress in clear, accessible language and acknowledge limitations. A useful monitoring system supports timely decisions and learning, rather than turning a legally defined target test into a complete judgment of every safety activity.