Ghost wages from proxy attendance and unverified headcounts quietly inflate labour cost. A practical look at how much it adds up to and how to catch it.
Attendance leakage is easy to overlook because it does not look like a single event — it is a small, repeated gap between who is marked present on the muster roll and who is actually on site working, and it accumulates quietly across a large labour force over a full wage cycle. Unlike a stock shortage or an equipment breakdown, there is rarely a single moment where the problem becomes visible unless someone is specifically looking for it.
"Buddy punching" refers to one worker marking, signing, or verbally confirming attendance on behalf of an absent colleague — the absent worker still gets paid for the day even though they were not on site. Ghost entries are a related but distinct problem: names remaining on the muster roll or wage list for workers who have left the project, are on extended unreported leave, or in some cases never existed as an actual worker on site. Both result in wages paid for hours not actually worked, but they arise from different gaps — buddy punching is a same-day verification failure, while ghost entries are typically a roll-maintenance failure that persists over a longer period.
A manual muster roll, maintained by a single supervisor or mistri without independent cross-check, has no mechanism to verify that the person marked present is the person physically on site. On a large site with multiple gangs and shifting daily headcounts, a supervisor recording attendance from memory or a verbal roll call at the start of the day has no practical way to catch a colleague covering for an absent worker, particularly if the absence is brief or intermittent rather than a full-day no-show.
A handful of proxy or ghost entries on any single day is a small percentage of a large labour force and rarely triggers concern on its own. But repeated daily across a full month, and across the life of a multi-month project, even a small daily leakage rate compounds into a material percentage of the total wage bill — commonly estimated in the 5-10% range on sites without any independent verification, though the actual figure varies by site and existing controls. Because it never appears as one large anomaly, it is one of the easier cost leaks to under-prioritise relative to more visible problems like material cost overruns.
The highest-value first step is independent cross-checking, not new hardware — having someone other than the person recording daily attendance conduct periodic physical headcount spot checks at varying times of the day, and comparing that against the muster roll. This alone catches a meaningful share of buddy punching, since it removes the assumption that the same person recording attendance is also the only person verifying it. Biometric or digital attendance systems address the problem more comprehensively and are worth considering for larger, longer-duration projects, but a basic cross-check process captures much of the available saving before that investment is justified.
Sites with tight attendance control separate the person recording daily muster from the person conducting periodic verification spot checks, and vary the timing of those checks so they cannot be anticipated — a predictable, always-morning headcount check is far less effective at catching proxy attendance than an unannounced check at a random point in the shift.
