Earned value management fundamentals

The cost report in Lesson 6 compares cost with budget line by line. It answers 'will this line overrun?' but it does not directly answer 'are we getting the work done we planned to, for the money we planned to spend?' Earned value management (EVM) answers that question with three numbers measured in the same units: what we planned to have done by now, what we have actually done, and what it actually cost. From those three numbers come variances and indices that summarize cost and schedule performance for a line, a control account or the whole project. This lesson defines the measures, works through them on Riverbend Commons, and explains where formal EVM systems are required.

Three measures

Earned value uses three cumulative quantities, all expressed in budget dollars (or budget hours) so that they can be compared directly.

Planned value comes from the cost-loaded schedule (Lesson 3). Earned value comes from measured percent complete (Lesson 4): for each item, EV equals percent complete multiplied by the item's budget. Actual cost comes from the job cost ledger with accruals (Lesson 5). The key idea is that EV converts physical progress into the same currency as the budget, so that it can be compared with both the plan and the actual cost.

Variances

Two variances compare earned value with the other two measures.

The cost variance (CV) is negative when the work done cost more than its budget: an overrun on the work performed. The schedule variance (SV) is negative when less work has been done than planned: the project is behind. Note that SV is measured in dollars, not days; Lesson 8 explains why that matters and how earned schedule converts it into time. In this course, following the usual EVM convention, negative variances are unfavorable.

Performance indices

Indices express performance as ratios, which makes it easier to compare control accounts or projects of different sizes.

The cost performance index (CPI) is the budget value earned for each dollar spent. A CPI of 0.95 means that each dollar spent has produced 95 cents of budgeted work. The schedule performance index (SPI) is the share of planned work actually done; an SPI of 0.90 means the project has done 90 % of the work planned by this date. Values above 1.0 are favorable; below 1.0 unfavorable.

A worked example at the project level

# Riverbend Commons, status at end of month 8 (amounts $)
BAC = 16_124_000
PV = 8_860_000        # from the cost-loaded schedule
EV = 7_907_000        # sum of percent complete x budget for every line
AC = 8_163_000        # job cost including accruals
CV, SV = EV - AC, EV - PV
CPI, SPI = EV / AC, EV / PV
print(f"Planned {PV / BAC:.1%} complete, actually {EV / BAC:.1%} complete")
print(f"CV = {CV:+,}  SV = {SV:+,}")
print(f"CPI = {CPI:.3f}  SPI = {SPI:.3f}")
print(f"Each dollar spent earned {CPI * 100:.1f} cents of budgeted work")

Planned 54.9% complete, actually 49.0% complete
CV = -256,000  SV = -953,000
CPI = 0.969  SPI = 0.892
Each dollar spent earned 96.9 cents of budgeted work

At month 8 the project had planned to be 54.9 % complete and is 49.0 % complete. It is behind schedule by $953,000 worth of work and has spent $256,000 more than the budgeted value of the work done. Both indices are below 1.0: cost performance slightly, schedule performance more. These are project-level figures; the useful next step is to see which control accounts produce them.

Drilling down

Project-level indices summarize, but they do not diagnose. A CPI of 0.969 could be a uniform small overrun or one large overrun hidden among many lines on budget. Earned value is therefore calculated for each control account and examined from the top down: first the project, then the accounts with the largest variances, then the items within them.

# Earned value by control account at month 8 (amounts $1000)
accounts = {  # name: (BAC, PV, EV, AC)
    "Site and foundations":       (1_420, 1_420, 1_420, 1_455),
    "Structure":                  (3_160, 3_160, 2_780, 2_960),
    "Envelope":                   (2_740, 1_644, 1_370, 1_390),
    "Mechanical and electrical":  (3_310, 1_324, 1_210, 1_200),
    "Interiors":                  (2_880,   360,   175,   180),
    "General conditions":         (1_904,   952,   952,   978),
}
print(f"{'Account':<28}{'CV':>7}{'SV':>7}{'CPI':>7}{'SPI':>7}")
for name, (bac, pv, ev, ac) in accounts.items():
    print(f"{name:<28}{ev - ac:>7,}{ev - pv:>7,}{ev / ac:>7.2f}{ev / pv:>7.2f}")
tot = [sum(v[i] for v in accounts.values()) for i in range(4)]
print(f"{'Project':<28}{tot[2] - tot[3]:>7,}{tot[2] - tot[1]:>7,}{tot[2] / tot[3]:>7.2f}{tot[2] / tot[1]:>7.2f}")

Account                          CV     SV    CPI    SPI
Site and foundations            -35      0   0.98   1.00
Structure                      -180   -380   0.94   0.88
Envelope                        -20   -274   0.99   0.83
Mechanical and electrical        10   -114   1.01   0.91
Interiors                        -5   -185   0.97   0.49
General conditions              -26      0   0.97   1.00
Project                        -256   -953   0.97   0.89

Structure produces most of both variances: it is $380,000 of work behind plan and $180,000 over on the work it has done. Envelope is behind but roughly on cost, which suggests it is waiting on the structure rather than struggling itself. General conditions are level of effort, so their SV is zero by definition (Lesson 4), but their CV of minus $26,000 shows they are running above the monthly budget. This pattern points the project manager at one place: the structure, and specifically the elevated slab forming that Lesson 5 already flagged.

Earned value in hours

For self-performed work, earned value is often tracked in hours rather than dollars, especially at the crew level. Planned hours, earned hours and actual hours give an hours CPI that equals the productivity factor of Lesson 5 (earned over actual) and an hours SPI. Working in hours strips out wage rates, so it measures productivity directly. Dollar-based EV is needed to combine labor with materials and subcontracts at the control account level; both are useful.

Conditions for meaningful earned value

Earned value is only as good as its inputs. Three conditions matter most. The baseline must be realistic and stable: if the cost-loaded schedule is front-loaded or not updated after approved changes, planned value is meaningless. Progress must be measured objectively by the rules of credit (Lesson 4); earned value calculated from optimistic percentages simply hides the problem one level down. And actual cost must be collected at the same level and for the same scope as the budget, with accruals, so that EV and AC describe the same work over the same period.

Formal earned value management systems

On most private building projects, earned value is an internal tool that the contractor applies as it sees fit. On some public and defense projects, the owner requires the contractor to run a formal earned value management system (EVMS) that meets a published standard, with reports in a specified format and periodic reviews by the owner.

In the United States, the industry standard is EIA-748, Earned Value Management Systems, maintained by the National Defense Industrial Association. A revision, EIA-748-E, was issued in February 2026; it reduces the guidelines from 32 to 27 by merging, deleting and adding guidelines while keeping five process categories. Federal civilian agencies apply EVMS under FAR Subpart 34.2 to major acquisitions as each agency sets thresholds. For the Department of Defense, a class deviation effective February 1, 2026 raised the threshold for EVMS reporting from $20 million to $50 million and kept a compliance review by the cognizant agency, usually the Defense Contract Management Agency, for contracts of $100 million or more. Construction contracts for federal facilities may also carry agency-specific cost and schedule reporting requirements, such as the US Army Corps of Engineers' cost-loaded schedule specifications.

Even where no formal system is required, the ideas of a formal EVMS are good practice: a defined scope broken into control accounts, a time-phased budget baseline, objective progress measurement, actual cost at the same level, variance analysis with thresholds, and disciplined control of baseline changes.

Earned value and the cost report together

Earned value and the line-by-line cost report are complementary. The cost report forecasts the final cost of each line from commitments and judgment; earned value summarizes performance to date and gives an independent, index-based forecast (Lesson 9). When the two disagree sharply, for example when the cost report shows a line on budget but its CPI is 0.85, someone's forecast is wrong and the difference should be explained. Used together, they make it hard for a problem to stay hidden.

Baseline changes and earned value

Earned value depends on a stable baseline, but the scope of a building project changes. When a change order is approved, its budget must be added to the baseline: to BAC, and to planned value in the periods when the change work is scheduled. If it is not, the change work will earn value against nothing, or its cost will appear as an overrun on base work. If the baseline is changed too freely, on the other hand, variances can be made to disappear by moving the plan to match the actuals, a practice sometimes called rubber baselining. Formal EVMS standards therefore restrict baseline changes to approved scope changes and a few defined cases, and require every change to be logged.