Introductory Econometrics - Chapter 1: The Nature of Econometrics and Economic Data

Flashcards covering introductory econometrics concepts from Jeffrey M. Wooldridge (2018), Chapter 1, including the definition of econometrics, steps in empirical analysis, economic vs. econometric models, the four data structures (cross-sectional, time series, pooled cross sections, panel data), and causal inference.

What is the formal definition of econometrics?

Econometrics is based upon the development and application of statistical methods for estimating economic relationships, testing economic theories, and evaluating and implementing government and business policy.

Why did econometrics evolve as a separate discipline from mathematical statistics?

Mathematical statistics focuses primarily on experimental data from controlled laboratory settings, whereas econometrics must deal with nonexperimental (observational or retrospective) data because controlled experiments on human behavior and economic systems are often impossible, prohibitively expensive, or morally repugnant.

What are the four primary steps in conducting an empirical economic analysis?

1. Precise formulation of the research question.
2. Constructing a conceptual model (a formal economic model or informal economic reasoning).
3. Specifying the econometric model.
4. Data collection and hypothesis testing using econometric methods.

How does an economic model differ from an econometric model?

An economic model consists of mathematical equations describing behavior (typically utility maximization under resource constraints), whereas an econometric model converts that abstract relationship into an estimable functional form that explicitly includes an error term ($u$) to capture unobserved factors.

How is criminal activity conceptualized in Gary Becker's (1968) economic model of crime?

It is conceptualized as a utility-maximizing resource allocation decision that weighs the economic rewards of illegal activity against opportunity costs (forgone legal employment) and direct costs (probability of arrest and conviction, severity of punishment).

In Becker's economic model of crime, $y = f(x_1, x_2, x_3, x_4, x_5, x_6, x_7)$, what do the variables $x_1$, $x_2$, and $x_6$ represent?

$x_1$ represents the wage for criminal activity, $x_2$ represents the hourly wage in legal employment, and $x_6$ represents the expected sentence length if convicted.

What is the error (or disturbance) term ($u$) in an econometric model, and why is it important?

The error term ($u$) represents unobserved factors, unmeasured variables, and measurement errors that influence the dependent variable; properly handling the error term is the single most critical component of econometric analysis.

In the econometric wage equation $\text{wage} = \beta_0 + \beta_1\text{educ} + \beta_2\text{exper} + \beta_3\text{training} + u$, what does the parameter $\beta_3$ measure?

$\beta_3$ measures the specific causal parameter of interest for job training on wages, holding education, experience, and unobserved factors fixed.

What is cross-sectional data, and what is its key ordering property?

Cross-sectional data consists of a sample of individuals, households, firms, cities, states, or countries collected at a given point in time (usually via random sampling); the ordering of the observations does not matter for econometric analysis.

What type of economic data structure is shown in Table 1.1 below?

Cross-sectional data, because it contains observations on 526 different individuals collected at a single point in time (1976), and the observation ordering is arbitrary.

What are the defining characteristics of time series data?

Observations are tracked chronologically over time, chronological ordering matters, successive observations are typically non-independent (serially correlated), and the data often exhibits trends and seasonality.

What type of economic data structure is presented in Table 1.3 below?

Time series data, because it tracks a single economic entity (Puerto Rico) chronologically across successive years from 1950 to 1987.

What is a pooled cross section, and what is its primary analytical purpose?

A dataset formed by pooling independent random samples drawn from the population at different points in time; its purpose is to increase sample size and analyze the effects of policy changes by comparing before-and-after periods.

What type of economic data structure is shown in Table 1.4 below, and what distinguishes it from panel data?

Pooled cross sections; it combines two independent samples of different houses sold across two separate years (1993 and 1995), rather than following the exact same houses over time.

What is panel (or longitudinal) data, and what is its primary advantage?

Panel data consists of a time series for each cross-sectional unit where the exact same units are followed repeatedly over time; its primary advantage is that it allows researchers to control for unobserved, time-constant characteristics (such as individual ability or city geography).

What type of data structure is illustrated in Table 1.5 below, and what are its dimensions ($N$, $T$, and total observations)?

A panel (or longitudinal) dataset, with $N = 150$ cross-sectional units (cities) tracked over $T = 2$ time periods (1986 and 1990), resulting in $N \times T = 300$ total observations.

What does the term ceteris paribus mean, and why is it central to econometrics?

It means 'other relevant factors being equal'; it is central because isolating the true causal effect of an explanatory variable $X$ on an outcome variable $Y$ requires holding all other confounding factors fixed.

Under the counterfactual (potential outcomes) framework, how is the causal treatment effect defined mathematically?

$te_i = y_i(1) - y_i(0)$, where $y_i(1)$ is the potential outcome if individual $i$ participates in the treatment and $y_i(0)$ is the potential outcome if they do not participate.

What is the Fundamental Problem of Causal Inference?

In real-world observational data, each economic unit can only ever be observed in one state of the world at any given time (either treated or untreated, but never both simultaneously).

In Example 1.3 (fertilizer on crop yield), why does randomized experimental assignment resolve the problem of causal inference?

Randomly assigning fertilizer amounts across identical plots ensures that unobserved factors (such as soil quality and parasites) are uncorrelated with fertilizer application, eliminating confounding bias.

In Example 1.4 (economic return to education), why does observational data make causal estimation difficult?

Individuals choose their own schooling levels, so unobserved characteristics such as inherent ability or motivation affect both education choices and wages, threatening to make any observed correlation between education and earnings spurious.

What does Example 1.7 (Testing the Expectations Hypothesis) illustrate regarding the use of econometrics?

It illustrates that econometric methods are used not only for causal inference, but also to test directly whether economic theories and market models (such as term structures of Treasury bill returns) hold true in real-world data.