Compound Growth and the Rule of 70

Why Compounding Matters

A constant percentage growth rate does not add the same amount to a level each period. It applies to the current level, so each period’s increase becomes part of the base for the next period. Over a long horizon, this compounding can make the difference between two growth paths much larger than a comparison of their annual rates might suggest.

This idea is useful when interpreting changes in real output or income measured over multiple years. The level and growth rate must be kept distinct: the level tells us how much output or income there is, while the growth rate describes the percentage change from one period to the next. When the quantity of interest is real, the growth rate should refer to a real measure rather than a change caused only by prices.

The calculations below assume a constant growth rate for clarity. Actual economic growth varies over time; a constant-rate path is a model for understanding accumulation, not a claim that real economies grow at an unchanging rate. This treatment is aligned with the SOA VEE Economics (Validation by Educational Experience) syllabus.

From a Constant Rate to a Compounded Level

Definition: Let $Y_0>0$ be a starting level and let $g$ be a constant growth rate per period, expressed as a decimal. If the level grows by the same percentage each period, then $Y_t$, the level after $t$ periods, is $$ Y_t=Y_0(1+g)^t. $$ For ordinary positive levels and growth factors, assume $g>-1$. A rate of $g=0.03$, for example, means growth of 3% per period. The model can be used with whole-number periods; a fractional $t$ is meaningful when the constant-rate formula is being used as a continuous interpolation of the discrete path.

The formula follows by applying the growth factor $1+g$ repeatedly. After one period, $Y_1=Y_0(1+g)$. After two periods, the new level is $Y_2=Y_1(1+g)=Y_0(1+g)^2$. Repeating the same step gives the formula for any whole-number $t$. The exponent counts how many times the growth factor is applied.

The total percentage change over $t$ periods is not $tg$ exactly. It is $\left((1+g)^t-1\right)\times100\%$. The product $tg$ can be a rough approximation when rates and horizons are small, but it omits the additional growth on earlier increases.

Worked Example: Projecting an Index at a Constant Rate

A hypothetical real-output index is 240 at the start of a five-year period. Suppose it grows by 3.5% each year, with the same annual rate applying to the level at the beginning of each year. Find the projected index after five years, the index-point increase, and the cumulative percentage increase. Round final answers to two decimal places.

Method. Apply the annual growth factor $1+g=1+0.035=1.035$ five times. This avoids the mistake of treating 3.5% as a fixed number of index points added each year.

The powers of the growth factor are $1.035^2=1.071225$, $1.035^3=1.108717875$, $1.035^4=1.147523000625$, and $1.035^5=1.187686305646875$. Therefore,

$$ Y_5=240(1.035)^5 =240(1.187686305646875) =285.04471335525. $$

The index-point increase is $285.04471335525-240=45.04471335525$, which rounds to 45.04 index points. The cumulative percentage increase is $(1.187686305646875-1)\times100\%=18.7686305646875\%$, which rounds to 18.77%. Thus the projected index is 285.04. As a check, adding the index-point increase to the starting level gives $240+45.04471335525=285.04471335525$, the same projected value.

Exact Doubling Time

A doubling time is the number of periods needed for a level to become twice its starting value. Under the constant discrete growth model, set $Y_t=2Y_0$ and solve for $t$. Because $Y_0>0$, it cancels from both sides:

2Y_0=Y_0(1+g)^t \quad\Longrightarrow\quad 2=(1+g)^t \quad\Longrightarrow\quad t=\frac{\ln 2}{\ln(1+g)}.

Result: For a positive starting level and a constant discrete growth rate $g>0$, the exact doubling time in periods is $$ t_2=\frac{\ln 2}{\ln(1+g)}. $$ This gives a positive future doubling time. If $g=0$, the level never doubles; if $-1<g<0$, the level declines, so it does not double at a nonnegative future time.

The logarithms work because they turn the unknown exponent into a multiplier: taking logs of $2=(1+g)^t$ gives $\ln 2=t\ln(1+g)$. The formula’s units are periods, matching the period used for $g$. An annual growth rate produces a doubling time in years; a monthly growth rate produces a doubling time in months.

Worked Example: Exact Doubling Time and the Rule of 70

A hypothetical regional income measure grows at a constant 2.5% per year. Estimate its doubling time using the Rule of 70, then calculate the exact doubling time under the discrete compounding model. State the time unit and round the exact result to two decimal places.

Method. The Rule of 70 divides 70 by the numerical annual percentage rate, so here the rate is entered as 2.5, not 0.025. For the exact calculation, use the logarithmic formula with $g=0.025$.

The Rule of 70 estimate is $70/2.5=28$ years. The exact calculation is $\ln(2)/\ln(1.025)$. Using $\ln(2)\approx0.69314718$ and $\ln(1.025)\approx0.02469261$,

t_2=\frac{0.69314718}{0.02469261} \approx28.0710\text{ years}.

The exact doubling time rounds to 28.07 years, compared with the 28-year Rule of 70 estimate. The estimate is close, but it is not the exact solution to the discrete model. A common mistake is to compute $70/0.025=2{,}800$ years. That incorrectly uses a decimal rate in a rule whose numerator is calibrated to a rate expressed in percentage points; $70/2.5$ is the consistent calculation.

The Rule of 70: A Useful Approximation

The Rule of 70 is a mental-math approximation for doubling time. Let $r$ be the numerical annual growth rate expressed in percentage points: for example, $r=2.5$ represents 2.5%, while the decimal rate in the compound formula is $g=0.025$. Then the rule is

t_2\text{ (approximately, in years)}\approx\frac{70}{r} \quad\text{or, using decimal }g,\quad t_2\approx\frac{0.70}{g}.

The approximation comes from the exact formula. For small $g$, $\ln(1+g)$ is approximately $g$, so $t_2=\ln(2)/\ln(1+g)$ is approximately $0.693/g$. Replacing 69.3 with 70 makes the calculation easy to do mentally. Since the rule rounds the constant and approximates the logarithm, it should be treated as an estimate rather than an identity.

For example, at 1% growth, the rule gives $70/1=70$ years; the exact value is about 69.66 years. At 5%, the rule gives 14 years, while the exact value is about 14.21 years. The usefulness of the estimate depends on the rate and the required precision. If a problem asks for an exact doubling time under a stated discrete model, use the logarithmic formula.

The rate must also match the time unit. If a level grows by 0.5% each month, the Rule of 70 gives an approximate doubling time of $70/0.5=140$ months, not 140 years. Converting that estimate gives about $140/12=11.67$ years. The annual rate is not 0.5% in this example; applying the rule directly to the monthly rate returns a result in months.

Why Small Rate Differences Become Large

Suppose two positive levels begin at the same value $Y_0$, but grow at different constant rates $g_1$ and $g_2$. After $t$ periods, their levels are $Y_0(1+g_1)^t$ and $Y_0(1+g_2)^t$. Their ratio is

\frac{Y_{2,t}}{Y_{1,t}} =\frac{Y_0(1+g_2)^t}{Y_0(1+g_1)^t} =\left(\frac{1+g_2}{1+g_1}\right)^t.

The starting level cancels, but the growth factors are still compounded for every period. If $g_2>g_1>-1$ and $t>0$, then $(1+g_2)/(1+g_1)>1$, so the second path ends above the first. Its percentage advantage relative to the first path is the ratio minus one, multiplied by 100%. This is different from the percentage-point difference between the annual growth rates.

Worked Example: Comparing Two Long-Run Income Paths

Two hypothetical regions each begin with a real income index of 30,000. One grows at a constant 2% per year and the other at 3% per year for 25 years. Calculate both final levels, their difference, and how much higher the faster-growing path is relative to the slower-growing path. Round final levels and the difference to the nearest cent, and the relative comparison to two decimal places.

Method. Compound each initial level using its own annual growth factor. Then compare levels in dollars and use their ratio for the relative percentage gap. The common mistake to avoid is to report the one-percentage-point annual rate difference as though it were the cumulative difference after 25 years.

For the 2% path, the 25-year factor is $1.02^{25}\approx1.640605994$. Its final level is $30{,}000(1.640605994)\approx49{,}218.17982$, or 49,218.18.

For the 3% path, the factor is $1.03^{25}\approx2.093777930$. Its final level is $30{,}000(2.093777930)\approx62{,}813.33790$, or 62,813.34.

Using the unrounded levels, the difference is approximately $62{,}813.33790-49{,}218.17982=13{,}595.15808$, or 13,595.16. The faster-growing path’s relative advantage is calculated using the slower-growing path as the denominator:

\left(\frac{1.03}{1.02}\right)^{25}-1 \approx0.2762223 =27.62223\%.

Thus the faster-growing path is about 27.62% higher after 25 years. The rounded levels provide a check: their ratio is approximately $62{,}813.34/49{,}218.18\approx1.27622$, consistent with a relative advantage of about 27.62%.

Assumptions, Interpretation, and Common Traps

Compounding calculations are only as meaningful as their inputs. Before calculating, identify the starting level, the rate per period, whether the rate is constant, and the number of periods. A growth rate must be paired with the period it describes; mixing an annual rate with a monthly horizon can produce a numerically neat but economically incorrect answer.

Adding rates instead of compounding. For a 3% annual rate over 10 years, the compound factor is $1.03^{10}$, not $1+10(0.03)$. The latter is a simple-growth approximation, not the constant-percentage model.
Confusing rate difference with cumulative gap. A 3% rate is one percentage point above a 2% rate, but that does not mean the final level is only 1% higher. The relative gap depends on how long the two growth factors compound.
Using the Rule of 70 with the wrong rate units. Enter 2.5 for 2.5%, or equivalently use $0.70/0.025$. Do not enter 0.025 into $70/r$.
Calling a decline a future doubling. The logarithmic expression does not represent a positive future doubling time when $-1<g<0$. In that case, the level shrinks each period.
Overstating what a constant rate means. The model describes a path under a fixed rate. It does not show that observed economic growth is smooth or constant from year to year.

A final interpretation check is helpful: a positive growth rate above zero makes the level rise, zero growth leaves it unchanged, and a rate between -100% and 0% makes it fall while remaining positive. Under a constant positive rate, the exact doubling formula and Rule of 70 apply to the stated rate and time unit; they do not automatically describe a changing sequence of rates.

Key takeaway: Constant percentage growth compounds as $Y_t=Y_0(1+g)^t$. For $g>0$, exact doubling time is $\ln(2)/\ln(1+g)$, while the Rule of 70 gives the quick estimate $70/r$ when $r$ is the numerical percentage rate. Long-run comparisons require compounding each path before comparing the resulting levels.

Check Your Understanding

For each problem, use the stated rate and period consistently. Answers involving money or levels are rounded as indicated.

A hypothetical output index begins at 500 and grows at a constant 4% per year for three years. Find the final index and cumulative percentage increase. Round the final index to two decimal places. (Answer: 562.43; 12.4864% cumulative increase)
A quantity grows at 1.8% per year. Estimate its doubling time with the Rule of 70 and calculate its exact doubling time under the discrete model. Round the exact result to two decimal places. (Answer: Rule of 70, about 38.89 years; exact, about 38.85 years)
At a constant annual growth rate of 5%, compare the Rule of 70 doubling-time estimate with the exact discrete doubling time. Round the exact result to two decimal places. (Answer: 14 years by the Rule of 70; about 14.21 years exactly)
A hypothetical positive index begins at 800 and declines by 3% each year for eight years. Find its final value and cumulative percentage decline. Round the final value to two decimal places and the decline to two decimal places. (Answer: 626.99; 21.63% decline)
Two paths start at the same positive level. One grows at 1.5% per year and the other at 2% per year for 30 years. How much higher is the faster-growing path relative to the slower-growing path? Round to two decimal places. (Answer: approximately 15.88% higher)

What You'll Learn