Steve Saville tells us:
The crux of the issue is that most of the gold mined in the past has not been consumed (burnt, eaten, used-up in industrial processes or built into structures); rather, it remains available today in readily saleable form. Consequently, gold-mine supply adds only about 1.5% per year to the total supply of gold, and changes in mine supply have very little influence on gold's supply/demand equation. . . .
By way of further explanation, imagine that gold were money. In this case, the contribution to total supply made by the mining industry would be the monetary inflation rate. If the inflation rate were zero, that is, if the gold-mining industry ceased to exist, then gold's purchasing power would tend to increase over the long-term at roughly the same rate as the economy grew. For example, if, during a 20-year period, there were no money (gold) supply growth and average annual economic growth of 3% then gold's purchasing power would be expected to gain a total of about 80% (3%/year compounded for 20 years) over the course of this period. At no stage would there be a shortage of money (gold), assuming that prices were permitted to freely adjust.