What is the welfare effect of a price change? This simple question is one of the most relevant and controversial questions in microeconomic theory and its different answers can lead to severe heterogeneity in empirical results. This paper returns to this question with the objective of providing a general framework for the use of theoretical contributions in empirical works, with a particular focus on poor people and poor countries. Welfare measures (such as Equivalent Variation or Consumer's Surplus) and computational methods (such as Taylor's approximations or the Vartia method) are compared to test how these choices result in different welfare measurement under different price shock scenarios. As a rule of thumb and irrespective of parameter choices, welfare measures converge to approximately the same result for price changes below 10 percent. Above this threshold, these measures start to diverge significantly. Budget shares play an important role in explaining such divergence, whereas the choice of demand system has a minor role. Under standard utility assumptions, the Laspeyers and Paasche variations are always the outer bounds of welfare estimates and consumer surplus is always the median estimate. The paper also introduces a new simple welfare approximation, clarifies the relation between Taylor's approximations and the income and substitution effects, and provides an example for treating nonlinear pricing. Stata codes for all computations are provided in annex.