How does salary typically increase over a career?

July 19, 2023 7:18 AM

I want to model a person's salary increase over the course of their career. I've built an initial model in which the person is assumed to receive the same percentage increase each year, but this feels wrong; I think that it should increase more quickly in early years and more slowly later. Are there existing rules of thumb or data curves that model this kind of thing? Assume white-collar office work.
posted by Beverley Westwood to Work & Money (17 answers total)

It depends on the company. Or more accurately, companies, plural. In practice, salary increases are typically fairly small year to year, with a larger bump every few years from moving to a new company.
posted by kevinbelt at 7:25 AM on July 19, 2023 [5 favorites]

I feel like most people have some big jumps (or periods of sharper increases) when they change careers, companies, or jobs. Sometimes someone gets a salary bump when they acquire a certification.

So yeah I don't think it's an even year-over-year increase for most people, I think it's a pretty bumpy ride.
posted by mskyle at 7:29 AM on July 19, 2023 [6 favorites]

Given the way corporate culture has operated over the last few decades, it might even be worth factoring in a certain chunk of people being shuffled into a lower paying role/laid off/forced into early retirement as part of calculations.
posted by DirtyOldTown at 7:36 AM on July 19, 2023 [4 favorites]

Yeah, cost of living increase almost every year, maybe a percentage point more if they've done a really good job, then only big jumps when they're promoted (junior, senior, principal, staff; or I, II, III, IV).
posted by Melismata at 7:37 AM on July 19, 2023

Agree. I'm a software developer, and I have years of low or no wage growth, punctuated by large changes when starting at new firms. Across four employers so far, I have had two meaningful promotions, total. The real increases were due to changing employers.
posted by Alterscape at 7:39 AM on July 19, 2023 [2 favorites]

Best answer: I would model a modest increase (3%?) within roles every year, and then maybe every 4-5 years a more significant bump (£5K?) - that's maybe either related to an internal promotion or (as is more likely for most of us these days) moving to a new role in another place.
posted by london explorer girl at 7:50 AM on July 19, 2023

I too received my most significant raises of my IT career by switching employers (3 times). My most recent (4th) switch was a pay cut, however, after a layoff (less pay for a less stressful job overall).

How Often Do People Change Jobs During a Lifetime? Did the pandemic change that data?
posted by Press Butt.on to Check at 7:57 AM on July 19, 2023 [1 favorite]

Sites like salary.com can be a good data source to start building this, but it'll vary a lot by industry too.

A few things to consider:
* An individual's income is going to probably have huge spikey jumps as they get promotions and move jobs; smeared out over a whole population, this will be much smoother year-to-year since nobody's coordinating their job changes with other people (ie, this year I get a new job, next year it's you, next year it's somebody else, but on average all of us change jobs every 3 years)
* The ratio of early to later-career salaries can vary a lot by industry. In tech, starting salaries are extremely high compared to most, for example, but then level off much faster than in other fields where there isn't as much room for "high paid individual contributors" and there's more of an expectation that a single mid- or late-career person manages a large number of junior people. In tech, a mid/late-career manager might make "only" 2x or 3x what a new graduate makes at the same company; my friends in advertising are more used to a 5x or more multiplier between junior and senior salary grades.
* In general, barriers to entry are going to matter a lot. The most obvious will be formal qualifications like certifications and degrees. Often you either can't get the job at all without them (leading to slightly higher initial salaries), or you start making a lot more once you get them. On the other hand, low barriers to entry mean it's easy to underpay at the entry level even if those who stick around later make a lot more.
* I don't know enough about this to model it, but company structures matter a lot too. For example, some fields are dominated by very large companies, leading to more long-term salary potential for those who climb the ladder. In other fields, it's more typical to have lots of smaller employers, which tends to mean salaries level out except for those who found their own (successful) companies at which point income is all over the place. This is related to how common freelance or consulting work is in that field. Profitability of a small business is not something that's easily modeled from averages; "just barely scraping by" is as plausible as "successful and expanding," which means that for fields with a lot of small companies, you'll have a wider range of later-career salaries than those dominated by very large employers with more-predictable structures.
posted by Tomorrowful at 8:09 AM on July 19, 2023 [2 favorites]

This is a "how long is piece of string?" type of question.

My father and I have both had very similar career trajectories as software engineers (over his 40+ years and my 20+ years as software engineers), so much so that any model you came up with would likely give us similar or identical outcomes. The only real difference is the companies themselves, and even then you'd be using hindsight from the current day to decide who would have made out better. We even once shared the exact same office suite once just at different times. We even were located in same building twice, so even "location of business" in your model wouldn't matter.

Anyway, this is probably "research that gets you a PhD in economics" sort of situation.
posted by Back At It Again At Krispy Kreme at 8:43 AM on July 19, 2023 [1 favorite]

Best answer: Krispy Kreme is correct: This is a subfield of labor economics; the search term you want is "age-earnings profile."

The specific profile you want depends on the specific question you are asking. If it's how average earnings change by age over the overall population, there's a handy data series here, produced by the Social Security Administration based on their earnings data. However, that series "reflects both the average earnings level of those who worked at that age and the percent of insured workers who actually worked at that age." So it declines at older ages to reflect the effect of early retirements on average ages.

Also, it combines the profiles of many different types of workers, which vary. For example, the age-earnings profile is generally steeper for workers with more formal education, so the profile for your office worker would probably be steeper than the aggregate profile.
posted by Mr.Know-it-some at 9:16 AM on July 19, 2023 [1 favorite]

Assuming this is for a long-term financial plan, you would also want to take inflation into account. I suggest building two variables into the model that you can tweak to see results. So you'd enter 2% inflation and 3% earnings growth, for example. For most people not on an executive track I would say that's a reasonably conservative combination, building in a COLA that equals inflation, plus 1% for assumed promotions over time.
posted by beagle at 9:47 AM on July 19, 2023

So like you I have an extensive financial model which covers the next 40 years, and I've built in assumption of 2% cash compensation growth per year when employed.

At the same time, I've tracked my year-over-year cash compensation growth every year for the last 20+, and it varies significantly:

40%
23%
16%
13%
14%
10%
-3%
-3%
23%
4%
6%
2%
37%
11%
14%
4%
3%
4%
16%
4%

For the sake of a financial model, you'll need to pick a number(s) that you're comfortable with, understanding that it will be wrong almost every year - that's the nature of a forecast. When it comes to growth estimates, I always go very conservative (2% cash compensation growth, 5% investment growth) but you need to decide what's right for you.
posted by NotMyselfRightNow at 10:00 AM on July 19, 2023

Step one to answering this adequately is understanding what you want to use this model for, and how much being wrong in either the short term or long term will matter. Because the variance in what a "typical" white collar worker (which encompasses so, so many fields of work!) makes over time is going to vary wildly based on all sorts of factors. Looking at a real life example, if I think about my graduating class and what they are all doing now (at least the ones still broadly in the same field), there's probably an order of magnitude difference in pay between the least well compensated and the most. So the model results based on the midpoint are not going to be super useful to people at the extremes, nor vice versa.
posted by Jobst at 10:30 AM on July 19, 2023

More than likely it's going to be bursty and you should not assume that it's going to be always upwards.

I've had a few years of small increases followed by big jumps. In 2005, I started a job that was about 50% more than I was making as a freelancer. Then I was offered a role that was a 50% jump over that in 2007. From there I was offered another raise to move into a new type of role.

That was stagnant for several years. Went back to freelancing and maintained the same level of income. Then I took another gig and that was a 40% increase.

The next job was flat two years later. Got a small bump two years after that, and another bump two years after that. Small increases and then in 2022 I went to another company and got another big bump.

That wasn't great so I took a new job at a major decrease. I expect minor increases the next year or two given the market.
posted by jzb at 10:33 AM on July 19, 2023

There was a great article about this in Psychology Today back in about 1978. It described how the pattern is different for dfifferent jobs and different industriies. Some start high and don't increase much, like engineering. Some start low, but pay very well if get to the top of the heap. Examples are education and journalism. Some start low and remain low. Etc.
posted by SemiSalt at 4:21 AM on July 20, 2023

One problem is that from your question it's hard to know whether you're modelling the average wage patterns, or a specific single person. And if the latter… well that will vary hugely.

Sticking with "office work", imagine a hypothetical person working in the public sector, not hugely ambitious, but gets a handful of small promotions until they reach the point that they don't want to, or can't, climb any higher on the ladder. And that's fine, and they work there till they retire.

At the other extreme, think of someone who ends their career as CEO of a huge company, earning millions in salary, and more in stocks, options, benefits, etc. That would be a very different pattern of pay increases over their career.

This sounds too obvious, but I don't think you can generalise, unless you first acknowledge that you're creating a model for a particular kind of person, career and industry, and then whether that's supposed to model the average, or some other point on the range of possibilities.
posted by fabius at 5:16 AM on July 20, 2023 [1 favorite]

Is this person male or female? White? Native English speakers? How swanky is this person compare to peers? Who do they work for.
posted by Lesser Shrew at 4:43 PM on July 21, 2023