How to compute lost labor hours with a measured mile, worked through a 12-week conduit example

A design change arrives in week five of a conduit run, and the electrical foreman’s crew slows down for the next two months. The project manager knows the change cost hours. The owner will want to know how many, and how they were counted.

One method for answering that is the measured mile. It compares the crew’s productivity on the same job during a period the disruption did not touch with its productivity during the period it did. The difference, multiplied by the quantity installed in the impacted period, gives the lost labor hours.

What the method compares

AACE International’s Recommended Practice 25R-03, the cost engineering society’s guide to estimating lost labor productivity in claims, defines productivity as output divided by input, such as linear feet of conduit per labor hour. It also defines a productivity factor, the actual rate divided by the baseline or planned rate.

AACE says there is no uniform agreement on how to calculate lost productivity, and also reports general agreement that a comparison to unimpacted work on the same project is preferred when enough data exists. AACE bases its practice on U.S. and Canadian case law and recommends legal advice before anyone prepares a lost productivity claim.

A paper written for AGC’s 2021 Surety Bonding and Construction Risk Management Conference describes the measured mile the same way. The authors, from The Vertex Companies, Watt, Tieder, Hoffar & Fitzgerald and Liberty Mutual Surety, say the method compares actual performance with actual performance, so it does not depend on whether the estimate was right.

They name two conditions:

1. A valid unimpacted period on the same scope of work. If the work was disrupted from its first day, there is no measured mile.
2. Quantities and hours recorded at the time the work happened, for both periods.

A worked example

The figures below are an example with round numbers. An electrical contractor estimated 50,000 linear feet (LF) of conduit at 0.110 labor hours per LF. In weeks one through four the crew installed 18,000 LF in 1,800 hours. That is 0.100 hours per LF, slightly better than the estimate, and it becomes the measured mile.

The owner issued a redesign at the end of week four. In weeks five through twelve the crew installed 32,000 LF in 4,160 hours, or 0.130 hours per LF.

The calculation has four steps:

1. Measured mile rate: 1,800 hours divided by 18,000 LF, or 0.100 hours per LF.
2. Hours the impacted quantity should have taken at that rate: 32,000 LF times 0.100, or 3,200 hours.
3. Lost hours: 4,160 actual hours minus 3,200, or 960 hours.
4. Cost: at an example loaded labor rate of $90 an hour, 960 hours come to $86,400 before markup.

The productivity factor for the impacted period is 0.100 divided by 0.130, or 0.77. In AACE’s terms, the crew worked at 77% of its own unimpacted productivity.

How it compares with other methods

The same job data gives three different answers depending on the baseline. The measured mile uses the crew’s own unimpacted rate. Comparing the impacted weeks to the estimate uses the bid rate. The total cost method takes all hours on the activity and subtracts the estimate.

The measured mile gives the largest figure here because the crew beat its estimate before the redesign. FindLaw’s review of a federal board decision makes the same point, that the method lets an efficient contractor use its own productivity rate, which can be better than the bid. The total cost figure is smallest because it also absorbs the 180 hours the crew saved in weeks one through four.

The AGC paper calls the total cost method the least favored approach and says earned value analysis, a comparison of hours earned against the estimate, requires showing that the estimate was realistic.

What a board accepted

In Clark Concrete Contractors, Inc. v. General Services Administration, GSBCA No. 14340, the General Services Administration Board of Contract Appeals accepted a measured mile on a cast-in-place concrete FBI field office in Washington. FindLaw’s summary of the decision reports that blast-resistance redesigns after the 1995 Oklahoma City bombing disrupted Clark’s concrete work.

Clark split the work into unimpacted, highly impacted and less impacted periods. For forming floor slabs, the reported rates were 0.048 labor hours per square foot unimpacted, 0.113 highly impacted and 0.065 less impacted. For stripping slabs they were 0.059, 0.147 and 0.108. Clark then ran the same steps as the conduit example for each period, pricing lost hours at its average labor rate.

As quoted in FindLaw’s summary, the board set the test this way: “We will accept a comparison if it is between kinds of work which are reasonably alike, such that the approximations it involves will be meaningful.” It rejected Clark’s attempts to improve the baseline by 10% for slabs and by 50% for stripping, because the record did not support either adjustment.

FindLaw reports a labor inefficiency award of $1,082,490 after overhead and G&A markups. The board’s decision could not be opened during this run, so these figures come from FindLaw’s summary.

How courts and boards have treated it

The AGC paper reprints a tally from Construction Schedule Delays by Dale and D’Onofrio (Thomson Reuters, 2018) covering 138 U.S. court and board cases. The measured mile was accepted in 16 of the 30 cases that referenced it.

In Turner Construction Co. (2017), a Smithsonian museum renovation, the AGC paper reports that two of three subcontractors using a measured mile proved their costs. A subcontractor that used an industry inefficiency factor lost because it did not produce the publication behind its 34% figure or expert testimony supporting it.

Records to keep from day one

The AGC authors say most contractors do not have project controls in place to track productivity, and that this view often changes after a damaging productivity impact. The measured mile needs the same daily inputs as labor productivity tracking: quantity by cost code from the daily log and hours by the same cost code from timecards.

The paper adds these practices:

1. Record daily quantities for each activity against the manpower on it, broken down by area.
2. When a crew splits its day across two tasks, code the hours to each task separately.
3. Build the measured mile early, while work is usually unimpacted, on activities that continue through the job.
4. Document each change to the planned work, such as stacking of trades, out-of-sequence work and access problems, with photographs and video.

A labor claim often sits beside other cost categories. Extended field overhead and unabsorbed home office overhead come from delay, and overtime premium comes from acceleration. Code the change orders by cause so each disruption can be tied to the period it affected.

To start, pick one repetitive self-performed activity on a current job, such as conduit or formwork. Compute its weekly hours per unit for the first four weeks and save that rate with the dates. If a disruption comes later, that saved rate is the starting point for a measured mile.