A reinforcing feedback loop (also called a positive feedback loop) is a loop in which a change in a stock is fed back in a way that amplifies the original change. More produces more; less produces less. The direction of change is self-reinforcing.
The label "positive" is often misunderstood: it does not mean good. It means that the feedback signal has the same sign as the change that created it. A savings account earning compound interest is a reinforcing loop — more money earns more interest, which adds more money. But so is a population of bacteria consuming nutrients, a panic in a bank, the spread of a virus through an unvaccinated population, and the accelerating loss of Arctic ice as melting exposes darker ocean that absorbs more heat and causes more melting.
Meadows, following Forrester, represented reinforcing loops with an R (or sometimes a + symbol circling the loop). The defining mathematical behavior is exponential-growth — or exponential decline. Exponential growth is characterized by a constant doubling time: a quantity growing at 2% per year doubles every 35 years, at 7% it doubles every 10 years. This was a central insight in limits-to-growth-1972: world population, industrial output, and resource consumption were all growing exponentially in a finite world.
The "snowball rolling downhill" is Meadows's intuitive image: a small initial advantage becomes self-reinforcing, gathering mass and speed. This dynamic creates virtuous cycles (early investors in a new technology gain advantages that attract more investment) and vicious cycles (poverty limits education limits earning power which deepens poverty; debt grows while an economy shrinks, making debt harder to service).
Reinforcing loops never operate in isolation for long. Eventually they encounter constraints — physical limits, resource exhaustion, competition, regulatory response — that activate balancing-feedback-loops to check the growth. The interaction between reinforcing growth loops and balancing limiting loops produces the characteristic S-shaped growth curves found throughout nature and social systems. When the balancing response is delayed or inadequate, overshoot-and-collapse results instead of a smooth leveling off.
In Meadows's leverage-points framework, strengthening or weakening the gain around a reinforcing loop is a moderately high leverage intervention (level 4). Identifying and dampening dangerous reinforcing loops — or, conversely, identifying and strengthening virtuous ones — is a key task of systems analysis.
The pedagogical challenge Meadows faced was that human intuition is poorly calibrated for exponential behavior. Linear extrapolation feels natural; exponential curves look like nothing is happening until they suddenly look like everything is happening. The doubling-time demonstration — that 30 doublings of any quantity starting from 1 produces over a billion — was a repeated teaching tool in her courses and columns.