In most econometrics textbooks the authors give an interpretation of a linear regression such as Y = a + bX, saying that a one-unit increase in X (years of education) will cause a b unit increase in Y (wages). Dealing with time-series regressions this may well be OK. The problem is that this ‘dynamic’ interpretation of b is standardly also given as the ‘explanation’ of the slope coefficient for cross-sectional data. But in that case, the only increase that can generally come in question is in the value of X (years of education) when going from individual to another individual in the population. If we are interested — as we usually are — in saying something about the dynamics of an individual’s wages and education, we have to look elsewhere (unless we assume cross-unit and
Topics:
Lars Pålsson Syll considers the following as important: Statistics & Econometrics
This could be interesting, too:
Lars Pålsson Syll writes The importance of ‘causal spread’
Lars Pålsson Syll writes Applied econometrics — a messy business
Lars Pålsson Syll writes Feynman’s trick (student stuff)
Lars Pålsson Syll writes Difference in Differences (student stuff)
In most econometrics textbooks the authors give an interpretation of a linear regression such as
Y = a + bX,
saying that a one-unit increase in X (years of education) will cause a b unit increase in Y (wages).
Dealing with time-series regressions this may well be OK. The problem is that this ‘dynamic’ interpretation of b is standardly also given as the ‘explanation’ of the slope coefficient for cross-sectional data. But in that case, the only increase that can generally come in question is in the value of X (years of education) when going from individual to another individual in the population. If we are interested — as we usually are — in saying something about the dynamics of an individual’s wages and education, we have to look elsewhere (unless we assume cross-unit and cross-time invariance, which, of course, would be utterly ridiculous from a perspective of relevance and realism).