In-house counsel are interested in how their compensation compares to others who are similarly situated. They might also be curious as to what most influences in-house compensation. Those factors, such as industry, revenue of employer, primary practice concentration, number of lawyers in the department and years since graduation from law school, are part of the 2012 General Counsel Metrics/Major, Lindsey & Africa In-House Counsel Compensation Report.

In this article we will explain the data and research underlying that report, with particular emphasis on the statistical tool of multiple linear regression, and some of the findings.

During 2012, General Counsel Metrics LLC and Major, Lindsey & Africa collected compensation data for approximately 1,800 in-house lawyers in Canada and the United States. Their combined set included base salary and cash bonus paid for fiscal year 2011, the year the lawyer graduated law school and the lawyer’s principal practice area of law. Along with this fundamental data, GC Metrics and Major, Lindsey & Africa collected corporate revenue, industry and departmental size data.

The full report includes 479 general counsel, but this article does not include them. Those officers of the company usually have significantly richer compensation arrangements than the rest of the lawyers. Nor does the lawyer information covered in this article include the value of equity awards because they are difficult to obtain and quantify. Accordingly the data summarized here are limited to 1,291 in-house lawyers in North America below the level of general counsel.

Aside from the usual graphics, charts and tables, analysts can investigate compensation data with statistics. One of the powerful statistical tools is multiple linear regression. Multiple regression shows how much changes in different "independent variables," such as years of experience or industry, influence a "dependent variable" — here, total compensation.

Think of multiple regression this way, starting with a single independent variable. For every additional year since a lawyer graduated from law school, what is the expected increase in that lawyer’s total compensation? In general, compensation rises with seniority, and that pattern shows up if you plot each lawyer’s total compensation against the lawyer’s year of law school graduation. The horizontal axis shows years out of law school, say from 1975 on the left to 2010 on the right. The vertical axis shows the total pay. You can see that the spray of data points generally slopes from the upper left down to the lower right.

For more precision, add to the scatter plot a "trend line."A trend line is the straight line that passes as closely as possible to all the data points of lawyer years versus total pay. More precisely, the line minimizes the sum of all the squared distances of the actual points from the calculated trend line. The trend line makes visual "regression" of years of practice on compensation.

Even more, the software that draws the line produces its formula. The slope of that line — how much compensation changes as lawyers practice longer — lets you do the calculation for any lawyer when you know the graduation year.

Now, add another potential determinant of compensation, such as the industry of the lawyer, or the size of the lawyer’s company measured by revenue. Each additional factor could produce its own scatter plot and trend line. One by one, it would be hard to see how all the factors play out. But multiple linear regression takes all the factors and determines their relative influence on total compensation. Thus, the regression is multiple. The more predictor variables you add, such as market capitalization, the lower all the other variables’ influence goes because each addition "explains" a bit of total compensation. The two numbers that jump out of any regression are "p" and "ß" (the Greek letter beta). ß is the estimated size of the linear effect; for example, it tells us how much compensation changes for each year the predictor variable changes. The "p" is how sure we are that the estimated size is exactly ß (a low p is better). The industry and practice area findings below had sufficiently low p values.

Statisticians use the term reliability to describe the confidence level they have in statistical results. Reliability increases with the size of the sample and with the concentration of data (less dispersion). For example, a confidence interval expresses in numbers how likely the linear formula is to the compensation of an actual lawyer. It takes quite a bit of compensation data to be able to make meaningful statements about subsets, such as tax lawyers in the retail industry.


This article focuses on the primary findings. As a framework, the median total compensation for this group was $187,195 with the first quartile at $143,000 (meaning that 75 percent made more than that) and the third quartile at $242,020.

• Years of practice. As you might expect, there is a strong positive correlation between number of years of practicing law and how much a lawyer is paid. In fact, multiple regression finds that to be the most significant predictor of compensation. On this data set, each additional year out of law school raised the expected pay by about $3,631. So, if you graduated in 1995, that would be 16 years of practice to 2011 multiplied by $3,654 equals $58,096 starting from the reference value. Here, the reference value is $134,310, so you would be likely to make something like $192,400.

• Revenue. It turns out, quite counter­intuitively, that the revenue of the company has almost no influence on lawyer compensation. One might think that larger companies can afford to pay more, but there is no significant correlation between larger company and larger paycheck. Noncash compensation may be different in bigger companies, such as pension plans, perquisites and equity, but cash compensation shows little effect.

• Industry. When we look at the industry in which a lawyer works, only two of them make much difference: extraction and medical devices. It is not clear why they stand out, but it could be the particular mixture of companies in this set. Otherwise, there were no statistically significant patterns. This finding seems odd because you might think that some industries would pay more. For example, those that are fast-moving and highly competitive — and need excellent legal talent — might pay more. On the other hand, the attraction of stock options and the allure of prominence may allow those flashy industries to pay less than more plodding, mature industries.

• Practice area. Within practice areas, the highest pay predicted by multiple regression belongs to regulatory special­ists, followed by securities, mergers and acquisitions, and practices specific to an industry, e.g., banking lawyers in banks, insurance specialists in insurance companies, and so forth. It makes sense that those lawyers who focus their practice on the legal issues of their industry might well have an edge when it comes to being paid. As for the other practice areas, it may be that compensation is relatively evenly distributed, varying more by years of experience or other factors.

Some compensation studies make claims like "Antitrust lawyers are the highest paid." That claim only makes sense if the analyst has controlled for industry and years out of law school, for example. Are antitrust lawyers really paid more just because they are antitrust lawyers, or is it because the particular sample of antitrust lawyers in the data set all happen to be very experienced? It’s possible to "explain" the observations (over the entire data set of all lawyers) more simply by positing a dependence on years experience.

It is certainly possible to analyze other factors that might influence the pay of in-house counsel. The ratio of lawyers per billion dollars of revenue, for example, might be insightful. Relatively fewer lawyers might tend toward relatively higher paid lawyers. The profitability of the industry might be a significant factor or whether the company is publicly traded or not.

Two final points deserve to be underscored: Managers in legal departments can make better decisions when they understand metrics and the tools available to interpret and present those metrics; and statistical tools are available to tease out the factors that are associated with different metrics, such as levels of pay. More broadly, multiple regression will come to play a prominent role in the analysis of in-house compensation and other law-department performance metrics. The GC Metrics/Major, Lindsey & Africa In-House Compensation Report exemplifies a sizeable data pool and powerful analytic tools.

Rees Morrison leads General Counsel Metrics, which offers law departments at no cost the world’s largest benchmark report, with more than 1,000 participants in 2012. Bob Graff is a partner and vice president – global business development for Major, Lindsey & Africa, an executive search firm for lawyers.