Periodically, general counsel wish that they had compensation data so they can compare what they make or what their lawyers make. If someone pushes hard for a raise or the department extends an offer to a new hire, it helps to know market rates. General counsel can buy such data from a few sources, find out pieces of such information online, or they can try to collect data from peer companies. Whatever the source, the data they typically obtain show base pay plus cash bonuses according to groups of lawyers with similar attributes, such as industry, years out of law school or primary practice area. The general counsel then draws conclusions from whatever data are available from the study.

Pages of tables of median amounts paid with little explanation are the common fare: Familiar and easy to read, tables pale in effectiveness against more sophisticated presentations. Much more could be done with compensation figures if there are enough of them and if the analysis and presentation move beyond the traditional format of tables. This article describes some of those better forms of presentation of, and therefore insights into, in-house compensation levels.

For a set of data to use here, I took compensation figures from lawyers below general counsel in about 60 law departments in one particular industry. Those law departments have participated in the General Counsel Metrics benchmark survey under way this year, and they provided their lawyers’ base in 2011, bonus, practice area and law-school graduation year as well as overall staffing and spending metrics for the department.

Visualize a scatter plot. On it, the horizontal axis along the bottom indicates number of years out of law school, one tick for each year starting at one on the left and extending right to 35 years post-graduation. On the left, add a vertical axis marked to show total cash compensation, starting at $50,000 in the lower left corner and moving up by $25,000 units to the top earner, at $350,000. Each of the 120 lawyers in this group becomes a point as high up the left axis as was their total cash compensation (salary plus bonus, but excluding equity awards and pension contributions, etc.) and as far over on the right as their years out of law school.

A lawyer 15 years out making $160,000 might be a point in the approximate middle of the square scatter plot. The graph looks like a sprinkling of points that mostly angles from the lower left toward the upper right — generally speaking, the longer a lawyer has been practicing, the higher the pay.

A statistical calculation to understand and quantify the relationship between years of practice and total compensation is known as linear regression. Spreadsheet programs can place a straight line within that cloud of dots such that the total distance between each of the dots and that line is the minimum.

A Formula

More importantly, the software determines the formula for that line. From my data, the linear regression of total compensation equals $107,532 plus $2,371 times the number of years out of law school. Thus, to figure out typical total compensation for a lawyer in this industry, start at $107,532 and multiply years since graduation from law school by $2,371, then add the two numbers. You can do the same if you know the components: base salary and cash bonus.

Telling even more, the software reveals how much of total compensation is predicted by years of practice. For this data, years out of law school only predicts about 50 percent of total compensation. Additionally, software can tell us how closely the data match the regression line. That number, known as the correlation of determination, was 0.666. This is a positive correlation, as pay rises with years of practice, but not a particularly strong one.

Finally, a statistical calculation called “Student-t” can tell us how well the data conform to a linear distribution. Many kinds of data do not and it is not a given that compensation data are distributed in an orderly pattern. My Student-t suggests a modest linear format. Clearly, according to all three measures there are other drivers of compensation, such as practice area and size of company. And all three inform us much more than tables can.

Imagine now similar scatter plots with a trend-line formula but this time for various practice areas. If “General Corporate,” “Litigation,” “Employment” and other legal practice areas are known, they offer a more precise understanding of how total compensation (or base or bonus) changes according to that crucial attribute: a lawyer’s specialty.

Consider, also, that usually the larger the company the higher the compensation packages for its lawyers. One way to portray and understand this link is with a bubble chart. The size of each lawyer’s point (such as pay vs. practice area in the last example) forms a bubble whose area is proportional to company revenue. Immediately you can see how the size of a company affects typical pay packages.

Another way to grasp the data is to create a box plot. A vertical box plot has a rectangle extending above each attribute on the horizontal axis, such as years out of law school. The top line of the rectangle is the third-quartile compensation for that year; the bottom the first quartile. If $200,000 is the third quartile, three-quarters of the lawyers made less than that amount. The median is a line within the box. Half of the lawyers made more than the median; half made less. Extending up from the third-quartile edge and down from the first quartile’s are dotted lines that convey the dispersion of compensation above or below the rectangle. Box plots usually extend those lines only to points 1.5 times the range between quartiles.

With a box plot for lawyers in a practice area or level, a general counsel can efficiently see how the particular incumbent in the department compares with others similarly situated. In all this, we assume the general counsel is looking at data for lawyers from the same industry.

The practice area of the lawyer also influences total compensation. This may have to do with the relative demand for lawyers in different industries. Intellectual property lawyers may, in general, be paid more in technology companies than insurance lawyers, for example. Mergers and acquisitions lawyers may get more than employment lawyers. How much more, on average, comes from yet another statistical calculation that tells whether the difference is “statistically significant” — should two close medians be deemed the same. For instance, $165,000 as a predicted total compensation might be plus or minus $6,000.

The Third Dimension

The contribution to total compensation of several attributes, conceivably including secondary practice area, advanced degrees obtained, quality of law school attended, LSAT scores and employees of the company, permits yet another statistical analysis: multiple regression. Whereas linear regressions as described above involve correlations between two variables, multiple regression can handle many variables. It determines how much each of those attributes contributes to the determination of total compensation. In other words, it weights those attributes and thereby gives a general counsel a much more nuanced understanding of someone’s pay as it compares with the lawyers in the comparator set. It may be that years with the company better predict someone’s compensation than anything else.

Our command of compensation data far exceeds what we can glean from tables of numbers when we graphically plot one attribute against another. But we can go further, and plot three attributes at once. In a three-dimensional plot, the left axis might be compensation, the horizontal axis years out of law school and the third axis — extending back in the cube — could be corporate revenue. Then, each lawyer would be located as a point in the cube having those three coordinates. It helps to admire one of these fascinating plots on a monitor where colors and different kinds of points (squares, triangles, stars) make the patterns clear. Superior to tables and cross-tabulations, a three-dimensional plot of significant attributes lets a general counsel match insightfully against a trio of attributes.

All of these statistical and graphical techniques to tell more about compensation are deployed by General Counsel Metrics. Scatter plots with trend-line formulas; box plots with key data compactly presented; multidimensional plots usefully coded; and multiple regression — all delve deeper into numbers and help a general counsel or human resources manager make more precise and defensible decisions. When coupled with a description of how to understand the graphs and calculations, as well as old-fashioned tables to go along, these 21st century advances are far more illuminating than traditional offerings.