Transcript cause
Lecture 1: What is Econometrics? BUEC 333 Summer 2009 Prof. Simon D. Woodcock What is Econometrics? Econometrics is the application of economic theory and statistical methods to analyze economic data. You learn plenty of economic theory in other classes, so here we'll focus on the statistical methods. It's both a science and an art. There's always a disconnect between economic theory and the real world. Creatively bridging this gap to apply economic theory to real-world data is an art. Economic data are rarely "perfect" (experimental) -sometimes requires some finesse to get believable results out of them. Economic theory and statistics are grounded in mathematics. Direct application of either is scientific. Empirical Questions, Empirical Answers Usually, we're interested in measuring or estimating some economic object. Hopefully, the thing we're estimating answers some interesting question. Examples: Does reducing class size improve the quality of education? What will be the value of the S&P 500 one year from today? Common sense tells us that students get more attention in smaller classes, and therefore probably learn more. But how much more? Reducing class size costs money (need more teachers). Does the improved quality of education justify the expense? We could answer these questions using data on test scores and class size: study the relationship between them to estimate how much test scores would improve if class sizes were reduced. Complications: what do test scores measure? What about neighborhood income/wealth? Do smaller classes really cause higher test scores, or are they correlated for some other reason? Knowing (with certainty) the answer to this question would be worth a lot of $. A good estimate is also valuable. This is a pure forecasting problem: estimate future value of S&P 500 based on its historical behaviour and relationship to “economic fundamentals” What's the elasticity of demand for a product? Manufacturers and retailers want to know this when setting price. Estimate this from consumption (e.g., scanner) data. What you’ll learn in BUEC 333 In this class, you will learn the statistical methods you would use to answer these questions (and many more ...) You will learn how to assess whether a particular estimate/analysis is any good. You will learn to use a specific statistical software package (EViews) You will get some practice working with real data. Why should you care? Knowing some basic econometrics can be very lucrative. Lots of jobs, some pay very well. Managers need to understand where estimates come from and what they tell you (and what they don't!) Helpful for understanding real-world statistics (polls, press about empirical studies, "facts" and arguments, etc.) Helpful for understanding finance (alphas, betas, R2, etc.) Knowing your way around some econometric software can too. Useful for future study (4th year courses -- you'll see/read quite a bit of empirical work, and maybe do some). Even if you end up doing something completely different, it's useful for understanding the world around you. Example: The Male-Female Wage Differential You often hear statements in the popular press like “In 1997, the average annual earnings of women working full-year, fulltime were 73% of men's. Using the average hourly wages of all employees, the ratio was 80%” (quote from StatsCan website) What does this mean? Is it interesting? Does it reflect discrimination? Or rational decision-making? Next ... We’ll begin a review of statistics: Things you should have learned in BUEC 232 but didn't (or that you forgot). Random variables Probability Probability distributions Expected Values Variance