ECO3203M Econometrics: Cross-sectional and Panel Data
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
Accountancy, Finance and Economics
Econometrics: Cross-sectional and Panel Data
ECO3203M
Part A: Data management (10 Marks)
1) Select and rename variables in Table 1. ‘Old name’ refers to the variable name in the original dataset while ‘new name’ is the new corresponding name to be defined. (5 Marks)
Table 1: Variable List
Survey Questions |
Old name |
New name |
GENERAL INFORMATION |
|
|
Year of the survey conducted |
year |
year |
Panel ID (the same ID for each firm across different years) |
panelid |
panelid |
What percentage of this firm is owned by Private foreign individuals, companies or organizations % |
b2b |
foreign |
What percentage of this firm is owned by Government/State % |
b2c |
soe |
BUSINESS-GOVERNMENT RELATIONS |
|
|
In a typical week over the last 12 months, what percentage of total senior management's time was spent in dealing with requirements imposed by government regulations? % |
j2 |
reg_time |
Over the last 12 months, has this establishment secured a government contract or attempted to secure a contract with the government? Yes/No |
j6a
|
gov_contract |
SALES |
|
|
During the past fiscal year, what were this establishment’s total annual sales? |
d2 |
sales |
During the past fiscal year, what percentage of this establishment’s sales were: Direct exports % |
d3c |
sales_exp |
INNOVATION |
|
|
During last fiscal year, did this establishment spend on formal research and development activities, either in-house or contracted with other companies, excluding market research surveys? Yes/No |
h8 |
formal_research |
During the last three years, has this establishment introduced new or significantly improved products or services? Yes/No |
h1 |
prod_innovation |
LABOUR and Capital |
|
|
Total number of permanent, full-time workers end of last fiscal year |
l1 |
employees |
During the past fiscal year, what was the net book value, that is the value of assets after depreciation, of the following: Machinery, vehicles, and equipment |
n6a |
capital |
2) Generate new variable exp_dum (export dummy) equals 1 if sales_exp is positive, otherwise 0. Generate new variable foreign_dum (foreign ownership dummy) equals 1 if foreign is positive, otherwise 0. Generate new variable soe_dum (state ownership dummy) equals 1 if soe is positive, otherwise 0. Generate new variable firmRD (firm R&D dummy) equals 1 if formal_research is 1, and otherwise 0. Generate new variable newprod (new product dummy) equal to 1 if prod_innovation is 1, and otherwise 0. Generate new variable labour_prod which is equal to sales divided by employees.
Produce a summary statistic of all variables in Table 1 and these newly created variables. Summary statistics table should include observation number, mean, standard deviation, minimum, and maximum values. (5 Marks)
Part B: Discrete Choice Variables (30 Marks)
1) Estimate the following equation with linear probability, Logit and Probit models, and compare the corresponding estimated coefficients obtained from these three different approaches. (7 Marks)
(1)
Where X is a set of explanatory and control variables, including foreign_dum, soe_dum, reg_time, and employees.
Based on the estimated coefficients obtained above, answer the following questions
2) Discuss what factors would potentially affect a sample firm’ likelihood to become an exporter? (10 Marks)
3) Holding other factors constant, discuss whether the likelihood of a foreign firm to become an exporter is significantly different from a non-foreign firm? (6 Marks)
4) Holding other factors constant, calculate by how much a firm’s chance to become an exporter would increase if reg_time grows from 20% to 30%? (7 Marks)
Part C: Instrumental Variable estimation (30 marks)
1) Innovation is an important factor in achieving productivity enhancement. The equation below models this hypothesis. However, innovation is regarded as endogenous, as firms that benefit from innovation tend to be more productive, while firms with a higher productivity level would choose to innovate more. Estimate equation (2) below using the OLS and TSLS. Use a logarithm of labour productivity as the dependent variable.
For TSLS, use newproduct as an instrumental variable. Compare results from the OLS and TSLS models regarding the variable of interest (firmRD).
(2)
(8 Marks)
2) Is the instrument valid? Is there evidence of a weak instrument problem? Explain and conduct appropriate test (7 Marks)
3) In the case when an instrument is not valid or is weak, what would be your suggestions to overcome these issues? (8 Marks)
4) Comment on the remaining threats to internal validity in the model. (7 Marks)
Part D: Panel Analyses (30 Marks)
Assuming total production is measured by Sales, and labour input is measured by the total number of employees, a simple Cobb Douglas production function can be written as:
(3)
Where A is the total factor productivity and Capital refers to the capital input. Taking logarithm on both side, equation (3) is transformed to
Cross-sectional: (4)
Panel: (5)
1) Using the same data you have collected, transform variables ‘Sales, Capital and labour (employees)’ into logarithm values. Estimate equation (4) with OLS and interpret the estimated coefficient. Discuss the capital and labour elasticities of sales, respectively. (8 Marks)
2) Based on variable ‘panel’, keep firms appearing at least two waves and drop the rest observations. Set the dataset in panel format, then obtain the estimated coefficients in equation (5) by using fixed effect approach. Compare the results with the one obtained from OLS and discuss which one is preferred. (8 Marks)
3) Include foreign_dum, soe_dum, reg_time, and sales_export as control variables in equation (4), then rerun the fixed effect model, discuss the findings. (8 Marks)
4) Using relevant economic theories or literatures, choose two additional variables from the questionnaire to add into the production equation (5). Rerun the new model and interpret the results. Based on the new findings, explain why these two additional variables are suggested (or not suggested) to include in equation (5). (6 Marks)
End of coursework questions. You should have answered all Questions.
Appendix: Country List
2015 |
2016 |
2017 |
2018 |
Indonesia* Ethiopia* Philippines* |
Cameroon |
Argentina |
Kenya** |
Côte d'Ivoire |
Bolivia |
Laos** |
|
Dominican Republic |
Colombia |
|
|
|
El Salvador |
Ecuador |
|
|
Honduras |
Guatemala |
|
|
Laos* |
Paraguay |
|
|
Nicaragua |
Peru |
|
|
Mali |
Liberia |
|
|
Myanmar |
Sierra Leone |
|
|
Zimbabwe |
Uruguay |
|
|
Niger** |
|
|
|
Togo** |
|
|
|
Benin** |
|
|
*: countries use idstd2015 for panelid
**: countries use n5a for n6a
2023-01-09