Linear growth increases by the same amount each step. Exponential growth increases by the same percentage (or ratio) each step. That single distinction creates wildly different outcomes over time: linear growth produces a straight line, while exponential growth produces a curve that starts slowly and then rockets upward.
The Core Math
A linear relationship follows the formula y = mx + b. Here, m is the constant amount added each step, and b is the starting value. If you earn $100 every day, your total after 10 days is $1,000, after 20 days $2,000, and after 100 days $10,000. The increase from one day to the next never changes.
An exponential relationship follows the formula y = a · b^x. Here, a is the starting value and b is the multiplier applied each step. If something doubles every step (b = 2), it goes 1, 2, 4, 8, 16, 32, 64, 128, 256, 512 in just ten steps. The increase from one step to the next gets larger every time because each new value is a percentage of an already-bigger number.
The simplest way to tell them apart in a table of numbers: linear data has equal differences between consecutive values, while exponential data has equal ratios between consecutive values. If each step adds 5, that’s linear. If each step multiplies by 1.5, that’s exponential.
What They Look Like on a Graph
A linear function always produces a straight line. The slope stays the same from left to right, which is why the word “linear” literally means “line-like.” An exponential function produces a curve. For growth (when the multiplier is greater than 1), the curve starts nearly flat and then bends sharply upward. For decay (when the multiplier is between 0 and 1), the curve drops steeply at first and then flattens out, approaching zero without ever quite reaching it.
One consequence that surprises people: in the early stages, a linear function can actually be ahead of an exponential one. If you’re adding 1,000 per step while another quantity doubles from 1, the linear quantity leads for the first several steps. But the exponential curve eventually catches up and then blows past it, because multiplying by a ratio compounds on itself. The crossover point depends on the specific numbers, but the overtaking is inevitable given enough time.
Money: Simple vs. Compound Interest
The clearest everyday example is interest. Simple interest grows linearly. You earn interest only on your original deposit, so the dollar amount added each year stays the same. If you put $5,000 into an account earning 5% simple interest for three years, you earn $5,000 × 0.05 × 3 = $750 total. Each year adds exactly $250.
Compound interest grows exponentially. You earn interest on your principal plus all previously accumulated interest. Invest $10,000 at 5% compounded annually for three years and you end up with $11,576.25. Of that $1,576.25 in earnings, $1,500 came from the principal and an extra $76.25 came from earning interest on interest. The gap between simple and compound interest widens dramatically over longer time horizons, which is why compound interest is often called the most powerful force in personal finance.
Biology: How Bacteria Multiply
Bacteria reproduce by splitting in two, a process called binary fission. One cell becomes two, two become four, four become eight. The population follows the equation N = N₀ × 2^n, where n is the number of generations. With a short generation time, sometimes as little as 20 minutes, a single bacterium can produce millions of descendants in under a day.
If bacteria instead grew linearly, adding, say, one new cell per generation, that same single bacterium would produce only a few dozen offspring in the same period. The exponential nature of doubling is what allows infections to escalate so quickly and why early treatment matters so much.
Why Exponential Growth Doesn’t Last Forever
Pure exponential growth requires unlimited resources, which don’t exist in the real world. Population ecologists describe what actually happens with a model called logistic growth. Early on, when resources are plentiful, a population grows exponentially. As food, space, or other necessities become scarce, growth slows. Eventually the population levels off at what’s called the carrying capacity of the environment. The resulting graph is S-shaped rather than the classic upward-shooting curve.
This pattern shows up far beyond biology. Technology trends follow it too. Moore’s Law, the observation that transistor density on computer chips doubles roughly every one to two years, held as a reliable exponential trend for decades. But recent analysis of Intel processor data from 1959 to 2013 found that the growth better fits a series of S-shaped curves, each with about 9.5 years of rapid growth followed by years of near-stagnation before a new technology phase begins. Today, chip manufacturers struggle to shrink transistors below 5 nanometers, and the pace predicted by Moore’s Law has clearly slowed. Exponential trends in the real world almost always hit a ceiling.
Humans Are Bad at Sensing the Difference
People consistently underestimate exponential growth. Researchers call this exponential growth bias, and it’s remarkably persistent. Even people who are aware of the bias still fall for it. The classic illustration is the legend of grains of rice on a chessboard: place one grain on the first square, two on the second, four on the third, and double each time. By the 64th square, you’d need more than 18 quintillion grains, enough rice to bury entire countries. Almost nobody intuits that number correctly.
The reason for the bias is that our intuition tends to think in straight lines. When we see something growing, we instinctively project the recent rate of increase forward as a constant addition rather than a constant multiplication. This matters in practical situations. During the early stages of an epidemic, for example, case counts double on a regular schedule, but most people mentally project a linear continuation of recent daily numbers. The result is that the eventual scale of the outbreak feels shocking, even though it was mathematically predictable from the start.
Computing: Linear vs. Exponential Time
In computer science, the distinction between linear and exponential determines whether a problem is solvable in practice. An algorithm that runs in linear time does an amount of work directly proportional to the input size. Double the input, double the time. Searching through a list one item at a time is a classic example.
An algorithm that runs in exponential time (often written as 2^n) does work that doubles with every single additional input element. Add one more item to the input and the computation time doubles. Add ten more items and the time increases by a factor of about 1,000. Add just 50 more and you’re waiting longer than the age of the universe. This is why problems classified as exponential-time are considered effectively unsolvable for large inputs, even on the fastest hardware available. The gap between linear and exponential isn’t just academic. It’s the boundary between “runs in a fraction of a second” and “never finishes.”
A Quick Way to Remember
- Linear: grows by adding the same number each step. Straight line. Predictable and steady.
- Exponential: grows by multiplying by the same factor each step. Curve. Slow at first, then overwhelmingly fast.
The difference comes down to addition versus multiplication. Over short periods or small numbers, they can look similar. Over longer stretches, exponential growth leaves linear growth so far behind that the two aren’t even comparable. That’s true whether you’re tracking money in a savings account, bacteria in a petri dish, or transistors on a chip.