Skip to main content
U.S. flag

An official website of the United States government

Official websites use .gov
A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS
A lock ( ) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.

Show me the incentive and I’ll show you the outcome

By Scott Kupor, Director, U.S. Office of Personnel Management

September 5, 2025

Charlie Munger wasn’t talking about the federal General Schedule (GS) pay scale when he addressed a 1995 Harvard crowd on the critical role that incentives play in driving employee behavior, but he may as well have been. Because, unfortunately, 30 years later, his premonition remains true: how we pay federal employees remains one of the most critical – and backward – drivers of employee behavior.

What is GS?

Congress passed the Classification Act of 1949 to standardize pay across federal agencies, resulting in the creation of the GS. Previously, agencies had been free to create their own pay schedules, which, not surprisingly, created high variability across the government.

The Classification Act set out to create greater uniformity in pay and – for that time period – was reasonably well designed. In the post-WWII era, most government employees were performing highly routinized, clerical roles – the age of the knowledge worker was still far in the future. As a result, a system that focused on the role description (rather than the performance of the individual assigned to the role) made sense. In fact, for most of these clerical roles there was little opportunity for individuals to distinguish themselves through differential performance; rather, grading of performance was largely pass or fail.       

Then – and today – the GS scale has 15 different levels (GS-1 being the lowest; GS-15 being the highest) for each federal job; the levels correspond to the functional responsibilities of the role. For example, a GS-9 program manager might do more basic project coordination functions, whereas a GS-15 program manager would be leading large-scale, complex programs. As a result, a new GS-9 is paid about $60,000 salary, whereas the GS-15 would earn about $150,000.

And within each GS-level, there are also ten “steps.” These are salary progressions within a level that enable an employee to increase her compensation prior to being promoted to the next GS-level. Using the above example, while our GS-9, Step 1 would earn $60,000, if she were ultimately to achieve a GS-9, Step 10 level, her compensation would be about $79,000. 

So, where’s the beef?

All sounds good, right? Employees get paid more as the demands/scope of the job increase, we eliminate major variability between federal agencies, and we accomplish the original goals of the Classification Act drafters – equality.

Not so fast. Going back to Munger, we need to look to the incentives the GS program creates to determine the behavior it instills in federal employees. And, as it turns out, the incentives are all wrong; equality should not trump performance.

We are not all the same

Simply put, the GS program does not recognize variability of performance or merit of an individual and thus incents mediocrity.

What if my colleague and I are both GS-9, Step 1 program managers and she runs circles around me in terms of her performance? Well, we can both get “promoted” to Step 2 after 1 year of government service, provided we each get at least a “Fully Successful” rating on our annual review (that corresponds to a score of “3” on a scale of 1-5).

And how many government employees generally receive at least a 3 on their annual review? Oh, about 99.7%! As long as I can keep myself out of the 0.3% of federal employees who receive a 1 or 2, I am treated exactly as my better-performing colleague. Seems like a bad incentive. 

Now, to be fair, there is an exception that managers can deploy – called the Quality Step Increase (QSI). By the way, note that just by its name – “Quality” – it conveys exactly my point; an employee’s quality differential relative to other employees is the exceptional way to increase their compensation. In my book, quality should be the norm, not the exception, for pay increases. And QSI is in fact exceptional – by far the vast majority of annual compensation increases for federal workers come from the regular step increase and cost-of-living adjustments.

But the QSI is fatally flawed as well. First, data shows only about 3-5% of GS employees get one each year. Second, all a QSI does is progress you to the next step in your same GS-level. Typically, that might be around a 3% compensation increase; it does not get you to a new GS-level altogether. Third, if I give you a QSI bump because you deserve it, I am prohibited from giving you another one – even if you continue to outperform – for at least one year!

So, my incentive is simple – work hard to get my QSI and then coast for 12 months since I am going to get another time-based bump just for being ranked in the middle of my class!    

Ultimately, the system rewards mediocrity – nearly 100% of federal employees will get a salary increase through a step adjustment each year simply for keeping the seat warm for 12 months. No other employer seeking to incent greatness would build a compensation systems based on tenure; merit-based performance hands-down is the better approach.

Don’t get me started

And that doesn’t even take into account the myriads of other issues with the GS system. 

A few quick examples:

  • Many of the GS levels have degree and “time in service” requirements. So, if I am a brilliant software developer who drops out of college and wants to devote my life to government service, my actual skills will not determine my compensation. Rather, the fact that I lack a college degree and have no prior work experience carries more weight than what my skills can contribute to the government. What a huge loss for our country!
  • And even within a specific GS level, I always start at Step 1, regardless of whether my skills are better than an existing employee who is at a Step 10 already. Again, equality and tenure trump performance.
  • While the government employs “locality pay” adjustments – recognizing the cost to hiring an engineer in San Francisco exceeds that in Des Moines– we fail to properly take into account job-type differentials caused by labor market dynamics. For example, if nurses are in short supply, we can’t just re-calibrate each of the nursing GS levels to allow government jobs to compete more effectively with private sector jobs. Dynamism based on an ever-changing job market is not in the government vernacular.

But all hope is not lost

We at the Office of Personnel Management (OPM) – with the support of President Trump – have a number of initiatives to try to Make Performance Great Again. True reform of the GS system will require congressional action – which we are going to pursue – but we are not sitting on our heels in the meantime.

On the front-end of candidate recruitment, we are executing on a full merit-hiring plan that will ensure candidates are evaluated based on actual functional assessments (vs self-attestation of their skills). This, of course, doesn’t address the GS scale, but will ensure we bring top talent into the government. 

We are also executing on the President’s performance management initiatives to address the 99.7% problem – it’s mathematically impossible that virtually every government employee is truly “fully successful” in their work product. By fixing this problem, we will create a high-performance culture that not only attracts great minds but will ensure that we keep them. And, even without reform to the GS system, fixing the annual employee ranking system will ensure that only truly top performers get merit-based rewards.

Finally, we are designing a number of new programs to bring in talent in critical areas – e.g., AI-related experts – that will rely on “exceptional” hiring and compensation authority vested in OPM. This will at least provide an opportunity to ensure compensation is based on merit contributions (not tenure or degree requirements) to expand the pool of talent for the government.

Sadly, what the then-Director of the OPM said in her 2002 report (which a lot of this post is drawn from) on federal pay remains unaddressed: “To attract the best and the brightest in this next generation into public service, we need a pay system that reflects the realities of the modern workforce where performance and results are emphasized and rewarded.”

With President Trump’s leadership, we will not let this problem go unsolved for another 23 years.

 

Control Panel