Quantitative Research Methods - Coursework Example

By using or combining several time series of cross-section observations, panel data provide “more informative data, more variability, less collinearity among variables, more degrees of freedom and efficiency” as per Baltagi’s discussion (1995, p. 3-6). 3. Panel data can study better the dynamics of change because the panel data technique can cover a repeated cross-section of observation across time. Thus, panel data is believed to be more appropriate than either of the two methods (cross-section and time series analysis) in the study of situations like successive waves of minimum wage increases across localities and local minimum wages and over time 4.

Panel data can detect and measure effects better than purely time series and cross-section data can do. 5. Panel data can study complicated behavioural models like economies of scale and technical change---better than what pure time series and pure cross-section analysis can do. For example, panel data can study the interaction of the variables involved over time, which cannot be done by pure time series data. 6. By covering more units over time, panel data can minimize the biases that may result as data are aggregated.

In 1995, Baltagi pointed out that time series and cross-section studies were not controlling for heterogeneity and run the risk of obtaining biased results (p. 3). For example, Baltagi (1995, p. 3) cited that consumption of cigarettes is often modelled as a function of lagged consumption, price, and income but the specification of the same function can vary across countries, state, and time. Baltagi 1995, p. 4) added that panel data can control for location-specific and time-invariant variables while a time-series study or a cross-section study cannot at that time.

As panel data can cover heterogeneity, Baltagi (1995, p. 4), not accounting for country heterogeneity can cause serious specification errors. In addition, Baltagi (1995, p. 4) said that panel data can study the dynamics of adjustment. Meanwhile, Verbeek (2008, p. 655) said that the main advantage of panel data over either time series or cross-section analysis is that through panel data, economists can specify more complicated and realistic models than a single time series or cross-section data can do. (b) Explain the intuition behind the fixed effect model (FEM) and describe the least square dummy variable (LSDV) and the time demeaned approaches to estimating a FEM.

[30 Marks] Verbeek (2008, p. 359) defined the fixed effects model as simply a regression model in which the intercept terms vary over the individual units. Gujarati (2004, p. 642) pointed out that the main intuition behind the fixed effect model or FEM is that although the intercept may differ across individual elements, each specific intercept does not change over time or is time-invariant. The methods for estimating the fixed effects model (FEM) are the least square dummy variable (LSDV) and the time demeaned approaches towards estimating the FEM.

In the LSDV method, the main instruments for capturing the fixed effects are dummy variables. The time demeaned variable approach to FEM modelling, reconstructs the basic model it = ? + ?it + it as departures of a variable from its mean over time or yit - I? (it - i ) + ( uit - )i) where the values with bars denote the time mean of the said variable.