Problems with Timekeeping and Payroll

Problems with Timekeeping and Payroll

Despite your best efforts to create an accurate timekeeping system, there are several types of errors that will arise from time to time and that require special controls to avoid. One is the charging of time to an incorrect job. This is an easy error to make, typically caused by incorrect data entry by a direct labor person, who, for example, transposes numbers or leaves out a digit. To keep this problem from arising, the timekeeping system can be made an interactive one that accesses a database of currently open jobs to see if an entered job number matches anything currently in use. If not, the entry is rejected at once, forcing the employee to reenter the information. This control can be made even more precise by altering the database to associate only particular employees with each job, so that only certain employees are allowed to charge time to specific jobs; however, this greater degree of precision requires additional data entry by the job scheduling staff, who must enter the employee numbers into the database for all people who are scheduled to work on a job. If there are many jobs running through a facility at one time, this extra data entry will not be worth the improvement in data accuracy. If the existing data entry system involves only a simple rekeying of data from a paper-based time card submitted by employees, the data must be interpreted and then entered by the data entry staff. But this generally results in the least accurate data of all, for now there are two people entering information (the employee and the data entry person), which creates two opportunities to make a mistake. In short, the best way to avoid charging time to the wrong job is to have an interactive data entry system.

Another problem is that a vastly inaccurate amount of hours will sometimes be charged to a job, usually through the incorrect recording of numbers. For example, an eight-hour shift might be entered incorrectly as 88 hours. To avoid such obvious mistakes, the timekeeping system can be altered to automatically reject any hours that clearly exceed normal boundaries, such as the number of hours in a shift or day. A more sophisticated approach is to have the timekeeping system automatically accumulate the number of hours already charged during the current shift by an employee, which yields an increasingly small number of hours that can still be worked through the remainder of the shift; any excess can either be rejected or require an override by a supervisor (indicating the presence of overtime being worked). This approach is not possible, however, if employees record their time on paper, since the information is entered after the fact, and any correction to an incorrect number will be a guess by the data entry person and hence may not be accurate.

Another possible problem is that an employee might charge an incorrect employee code to a job, resulting in the correct number of hours being charged to the job but at the labor rate for the employee whose number was used, rather than the rate of the person actually doing the work. To avoid this error, the timekeeping system should be set up to automatically access a list of valid employee numbers to at least ensure that any employee code entered corresponds to a currently employed person. Though this is a weak control point, it at least ensures that hours charged to a job will be multiplied by the hourly labor rate of someone, rather than by zero. A much stronger control is to require employees to use a bar-coded or magnetically encoded employee number that they carry with them on a card, which ensures that they enter the same employee code every time. A weaker control is to post a list of bar-coded or magnetically encoded employee numbers next to each data entry station—it is weaker because an employee can still scan someone else's code into the terminal. If a paper-based system is used, an employee normally writes his or her employee number at the top of a time report, which is then entered by a data entry person into the computer at a later date. The problem with this approach is that the data entry person may enter the employee number incorrectly, which will charge all of the data on the entire time report to the wrong employee number. Again, an interactive timekeeping system is crucial for the correct entry of information.

Yet another problem is that the cost per hour that is used by the timekeeping system may not be the same one used in the payroll system. This problem arises when there is no direct interface between the timekeeping and payroll systems, meaning that costs per hour are only occasionally (and manually) transferred from the payroll system to the timekeeping system. This results in costs per hour on timekeeping reports that are generally too low (on the grounds that employees generally receive pay increases, rather than decreases, so that any lags in data entry will result in costs per hour that are too low). One way to fix this problem is to create an automated interface between the payroll and timekeeping systems, so that all pay changes are immediately reflected in any timekeeping reports that track labor costs. It is important that this interface be fully automated, rather than one that requires operator intervention, otherwise there is still a strong chance that the cost data in the timekeeping system will not be updated, due to operator inattention.

An alternative approach is to keep all labor costs strictly confined within the payroll system and to import timekeeping data into it, rather than exporting payroll data to the timekeeping system. There are two reasons for taking this approach: First, exporting payroll data anywhere else in a company makes it easier for unauthorized employees to see confidential payroll information; second, the payroll system cannot generate many meaningful reports without data from the timekeeping system, whereas the timekeeping system can generate a number of reports that do not need labor cost data (see the earlier section on timekeeping reports). Thus, it may be better to leave the payroll data where it is and instead work on an automated interface that imports timekeeping data into the payroll system.

Not only is it entirely possible that any of the problems described in this section will occur, but it is also possible that they will go undetected for a substantial period of time. To avoid this happening, the internal auditing department should be asked to conduct a periodic review of the controls surrounding the timekeeping and payroll systems, as well as a test of transactions to see if any problems can be spotted. The resulting audit report can be used to further tighten the controls around these data collection systems.

In the Real World: Reducing the Cost of Timekeeping

A routine analysis of the system costs at a large manufacturing facility discovered that the cost of administering the company's direct labor timekeeping system appeared to be inordinately high. Approximately 50 percent of the entire cost accounting function was devoted to the collection and interpretation of data related to direct labor. The controller asked a cost accountant, Ms. Anna North, to investigate the situation and recommend a revised system that would generate usable information, while costing as little as possible to administer.

The cost accountant's plan for this analysis was to, first, determine the level of detailed information collected by the timekeeping system, second, calculate the cost of collecting it, and then determine the benefit of using the resulting information. She would then see if costs could be reduced for the existing collection system, while losing no benefits from the system. If this was not possible, or if the costs could be reduced only by a modest amount, then she would investigate the possibility of reducing the level of information gathered, which in turn would reduce the cost of data collection.

Ms. North's first step was to determine the level of detailed information collected by the timekeeping system. She interviewed the facility's controller, Ms. Barbara MacCauley, who said that the timekeeping system required employees to write down on a time sheet the hours they worked each day on specific jobs, as well as the workstations where they worked on each job. The typical time sheet looked like the one shown in Exhibit 2.5.

Employee Name: Mort Dulspice

Date of Time Card: 4/13/03

Time In	Time Out	Job	Work Center
08:00	08:45	004712	Lithograph
08:46	09:12	004712	Etching
09:13	10:48	004712	Lamination
10:49	12:00	004712	Glue
01:00	02:10	004799	Lithograph
02:11	03:04	004799	Etching
03:05	03:17	004799	Lamination
03:18	04:24	004799	Glue
04:25	05:00	004799	Packaging

Exhibit 2.5: Atlanta Facility Time Sheet

It was apparent from the time sheet that each employee must carefully enter a large amount of information during the course of a shift. Also, the information entered by the employee in the example was not easy to read, making it likely that the person who entered this information into the computer would have a difficult time doing so correctly. Further, many time sheets were submitted each day by the 412 direct labor personnel at the facility, some of which were lost by employees or during the data entry process. This information had to be re-created, which could only be done through estimates of the work an employee completed during the period.

Ms. North found that these three issues gave rise to three different types of costs. The first cost was the time required by employees to enter their time worked onto each time sheet and then transport that time sheet to the payroll office for data entry. The second cost was for the data entry staff to initially enter the data into the computer; the third cost was to track down and correct any missing information or to correct data that was inaccurately entered. Ms. North calculated these costs for a typical month in the following manner:

1. Cost to initially record data. She estimated that each employee required 10 minutes per day to complete and deliver his or her time sheets. Since the average burdened cost per hour for all 412 employees was $17.92, this resulted in a monthly cost of $25,869 to collect the information, assuming 21 business days per month (412 employees x 21 days x $2.99/day).

2. Cost to enter data into computer system. She found that one and a half employees were required in the accounting department on a full-time basis to enter into the computer system the information from all 412 time sheets. These hourly employees earned a burdened wage of $12.05 per hour. This resulted in a monthly cost of $3,037 (1.5 employees x 21 days x 8 hours/day x $12.05/hour.)

3. Cost to correct data errors. On average, the accounting staff spent three hours per day correcting errors that had been discovered on time sheets or created during data entry. These errors were investigated and corrected by a senior data entry clerk, whose hourly burdened pay was $15.28. This resulted in a monthly cost of $963 (3 hours x 21 days x $15.28).

The grand total of all these costs was $29,869 per month, or $358,428 per year.

Ms. North's next task, determining the value of the benefits derived from the timekeeping system, was much more difficult. She found that the number of daily hours worked was used by the payroll staff to calculate and pay weekly wages to the direct labor employees. She described this function as a mandatory one for which the system had to provide sufficient data to calculate the payroll, but she could not ascribe to it a monetary value.

Next she looked at the benefit of tracking hours by job worked. This information was used by the cost accounting staff to develop an income statement for each job, which the sales staff used to revise its pricing estimates for future jobs, to verify that pricing levels were sufficiently high to ensure a targeted profitability level per job of 30 percent. The proportion of direct labor to all job costs was about one-third, so this was considered a significant cost that must be tracked for this purpose. The pricing staff members assured the cost accountant that they frequently altered their pricing strategies in accordance with the information they received through the job income statements. Once again, Ms. North found herself unable to clearly quantify a benefit associated with the tracking of direct labor hours, this time in relation to job numbers, but it appeared that obtaining the information was mandatory.

Ms. North's last benefits-related task was to quantify the benefit of tracking labor hours by workstation within each job. She found that this information was only used by the industrial engineering staff, whose members summarized the information into a report that listed the total hours worked at each workstation, by day, so that they could determine when capacity utilization levels were reaching such heights that new equipment had to be purchased or when levels were so low that existing equipment could be sold. A brief discussion with the production scheduling staff revealed that standard capacity amounts per job were already stored in the labor routings of the facility's manufacturing resources planning (MRP II) system, which produced a similar report by multiplying the units in the production schedule by the hours per unit of production listed in the labor routings. This meant that an alternative system could be used to provide the industrial engineering staff with the information it needed, without resorting to additional data entry to provide this information.

Ms. North then perused sample time sheets submitted by employees and noted that an average of three workstations were referenced on each time sheet for each job on which work was performed. If she could convince the management staff to eliminate the tracking of time by workstation, she could cut the labor time spent by the direct labor employees on timekeeping by two-thirds, plus similar amounts by the data entry clerks who would otherwise have to enter and correct this information, since these additional entries would no longer have to be made. This worked out to a cost savings of $19,912 per month ($29,869 x 2/3), or $238,950 per year.

Ms. North realized that the industrial engineering staff would agree to this change only if she could prove that the data the staff received from the MRP II system was sufficiently accurate to replace the workstation capacity data it had been receiving from the timekeeping system. To ensure that the MRP II system maintained a high level of labor-routing accuracy (which was the prime driver of the accuracy of capacity information produced by the MRP II system), she added $50,000 back to her estimate of remaining timekeeping system costs, which would pay for an engineer whose sole purpose was to continually review the accuracy of labor routings. This resulted in a timekeeping system cost of $169,478, which still represented a reduction of $188,950 from the earlier timekeeping system, for a drop in costs of 53 percent.