Black–Scholes/Binomial Lattice. These models are used to determine a poten- tial value for stock options. About 80 percent of companies use the Black–Scholes methodology; survey providers should display the methodology they use in valuing stock options.
Equity. This term is one of the more difficult compensation elements to survey.
While base salaries and incentives use a common “language” (dollars, in the United States), there are various equity vehicles as well as various ways to quan- tify the size of equity grants. A complete discussion of this topic is beyond the scope of this chapter, but following is a brief discussion of the key issues:
complete picture of the equity competitive market, certain assumptions need to be made regarding relative equivalencies, and such assumptions need to be clearly stated in the survey report.
■ Types of grants—equity grants include those made at or about the date of hire, grants made periodically after hire (on an annual basis for some position levels, sometimes referred to as “ongoing grants”) and those made on an unscheduled basis or for special reasons, such as promotion grants, retention grants, grants in connection with a merger or acquisition, etc. It is important that the survey presents data on different types of grants separately so that survey users can evaluate the competitiveness of specific features of their own programs. For example, the size of new hire grants is often a multiple of the size of annual ongoing grants. Combining these two types of grants would dilute the effectiveness of using the survey data for either situation.
Some surveys further segregate equity grants reporting into grant guidelines and actual grants. The former is the equity equivalent of “target incentives,” and can be useful in developing a picture of typical equity practices in the market.
However, not all companies have guidelines, and those that have guidelines do not always follow them, so it is also important that the survey captures and reports equity grants that have actually been made.
■ Equity quantification approaches—there are various ways to quantify the size of an equity grant, ranging from number of shares or percentage of company outstanding shares (which do not attempt to “value” the grant) to starting value or face value (the number of shares multiplied by the stock price at date of grant) to net present value (NPV) or Black–Scholes value, which include the potential future value of the grant, based upon certain assump- tions. Different types of equity vehicles lend themselves to one or more of these quantification approaches, and some more comprehensive surveys will report equity data using several different quantification methods. For example, stock option amounts are often reported in terms of NPV or Black–Scholes value, as well as face value or number of shares, while restricted stock amounts are typically reported in terms of face value or number of shares.
■ Combining equity vehicles—while it would not be appropriate for a survey to combine new hire grants with ongoing grants (as noted above), it is useful to combine the types of equity vehicles commonly used in the market to provide a more complete picture of the overall value of equity provided in a competi- tive environment. For example, in the wake of accounting rule changes requiring stock option expensing, the use of restricted stock has become more widespread, often used either in tandem with stock options or in
SALARYSURVEYS 121
lieu of options. A survey that presents data for options and restricted stock combined, as well as separately, provides a more complete view of the mar- ket. However, the value of a stock option and a restricted share are not equal, so certain assumptions need to be made when combining these together. For example, some surveys will combine the “calculated value” of options (using either NPV or Black–Scholes) with the face value of restricted stock to arrive at a combined value. Others may provide a combined number of shares that restates restricted stock in option equivalents, based upon certain conversion ratios. In either case, the survey vendor should clearly state the methodology and assumptions being used. A survey user can then rely on the same methodology and assumptions to quantify his company’s own equity grants and make an “apples-to-apples” comparison to the survey data.
International Surveys. Many companies have employees in more than one country and require data on local competitive practices. This can present a partic- ular challenge when surveys cover different compensation elements across countries, have varying definitions of what is included in a given element, and use different benchmark jobs and leveling schemes. Surveys in each country need to capture all the components of compensation, such as car allowances, housing, meal and transportation allowances, extra months’ pay, holiday or festival bonuses, etc. It is important to present the data in a consistent format, suitable for both local and global users. As multinational companies move toward global job leveling, where job X is at the same level from country to country (although pay levels may be different), surveys with consistent global platforms are becoming essential tools in compensation planning.
Mean or Average. The mean is the average of a set of values, calculated by adding all the numbers and dividing by the number of values. Contrary to the median, averages are more susceptible to the influence of outliers.
Median. The median describes the midpoint of a range of numbers; half the numbers in a range are above the median, and half below. When there is an even number of numbers in the range, the middle two are added and divided by two to calculate the median. The median is used in working with compensation data to mitigate the impact of pay that is at the extreme highs and lows of the range. It is not unusual to see significant variability between the highs and lows of compen- sation data, as actual pay can be affected by several factors, including length of service of the incumbent, variability in the responsibility or criticality of the role,
the 25th, 50th, and 75th percentiles; some surveys will include the 10th and 90th percentiles to provide a more complete picture of the range of the data.
Weighted and Simple Average. The weighted average (sometimes also called the “employee” average) takes into account all incumbents in the data sample.
Companies matching more employees to a survey job will have a greater impact on the survey weighted average than companies reporting fewer employees.
Weighted averages are used in compensation surveys to express the overall market rate for a particular job. On the other hand, a simple (or “company”) average weights the data for each company in the sample equally and is essentially an aver- age of each company’s average for the data element. Simple averages can be use- ful when a small number of companies dominate the data sample.
SUMMARY
Salary survey data is always a snapshot of a moment in time, and the application of the data is part art and part science. As labor markets are a consistently chang- ing picture, the limits of even the best pay data should be well understood. Good data leads to more great questions than simple answers. It is tempting to use salary data to find “one number” to the question of “how much should we pay?” or “how should we pay X employee?” However, because the market is always changing, the market will always be defined as a range of values and not a discrete number.
Determining the appropriate range—and choosing a value within that range for a specific recommendation—requires a combination of the “science” found in the calculated values and the “art” of using these figures in alignment with an organi- zation’s strategy and pay philosophy.
SALARYSURVEYS 123
10
125
C H A P T E R