Dear Prafulla Nath,

in order to model the relationship you adduced, the VAR could be a good choice, but under some conditions. It is preferable that your dependent and explanatory variables are stationary at first differences; that implies that you conduct some of the unit root tests. The time span you analyze is limited by data availability, it is better to use longer time series, or focus on time periods relevant to your research problem. The ideal lag depends on the data you are about to analyze. You should consult different information criteria in order to make an appropriate choice of the lag length. An alternative approach to your analysis could be an Autoregressive Distributive Lag (ARDL) since it allows modeling the relations between variables regardless of their order of integration (pay attention that none of them should be I(2) which is indesirable).

I hope that those suggestions will help.

Regards,

Vladimir

Dear Vladmir,

Thank you for your kind suggestion.

Prafulla

Dear Prafull,

Vector autoregression (VAR) is a stochastic process model used to capture the linear interdependencies among multiple time series. VAR models generalize the univariate autoregressive model (AR model) by allowing for more than one evolving variable.

Forecasts from VAR models are quite flexible because they can be made conditional on the potential future paths of specified variables in the model. In addition to data description and forecasting, the VAR model is also used for structural inference and policy analysis.

In health economics, and many other areas of applied economics, we face problems of endogeneity amongst dependent and independent variables. VAR model performed well, and so the technique has become popular in economics despite critiques that it is atheorectical. The basic idea behind the model is to treat all variables symmetrically, such that variables which that we are not confident are exogenous are modelled as endogenous. This leads to an n-equation, n-variable linear model, where each variable is explained by its own lagged values, plus the current and past values of the other lagged variables. While VAR models are often used in macroeconomics to analyse the relationship between different policy tools, they have rarely been used in the area of health economics. VAR models can be very useful for data description.

Hospital performance over the period 1996 to 2008 is evaluated by using VAR model. The regressions are run separately for each year (t) and outcome measure (k), and the hospital intercepts, representing the mean value of outcomes of each hospital holding patient characteristics constant across all hospitals, are extracted and used to create a new dataset at the hospital level. this model can be used for population labor force estimation, female, fertility…all regression you could have done using this approach. A Vector autoregressive (VAR) model is useful when one is interested in predicting multiple time series variables using a single model.

Y k iht = βqk 1h + "φXjht + uiht .

Please go through the cite I mentioned here.

http://eprints.lse.ac.uk/36674/1/LSEH_WP22.pdf

The data available in the HES database contains patient characteristic data (e.g. gender, age), clinical information (e.g. diagnoses, procedures undergone), mode of admission (emergency, elective), outcome data (mortality, readmission, discharge location) as well as details on the amount of time spent in contact with the health system (waiting times, date of admission, date of discharge) and details of which hospital the patient was treated in.

Ashish

Thank you prof. Ashis for your detailed information. I will go through your suggested materials.

Prafulla