Percentages can be highly misleading because they abstract from the hard reality of absolute numbers. This is especially true when considering the exponential nature of population growth.
The cited percentage change would be far more encouraging if population was constant, but it's not.
That graph is more informative. In terms of absolute numbers, extreme poverty has decreased by about 30% since 1820.
I tried finding a similar graph/data set in regard to "regular" poverty (rather than "extreme") but couldn't find one. Does anyone have a source for the change in absolute numbers of regular poverty?
The cited percentage change would be far more encouraging if population was constant, but it's not.