联系方式

  • QQ:99515681
  • 邮箱:99515681@qq.com
  • 工作时间:8:00-23:00
  • 微信:codinghelp2

您当前位置:首页 >> javajava

日期:2019-11-16 09:17


GradedAssignment

Please submit all answers in one word document. Where appropriate copy and paste

STATA output and code into this document.

Please submit your answers online via the Hub before 12:00 on Friday, September 29,

2017.

Question1 (20 points)

Consider the file brlongmerge.dta. It contains various statistics at the level of 380 regional

authorities of the UK over 10 years. The variable regid is a unique regional identifier.

(a) The variable domCO2Opop contains the domestic CO2 emissions (i.e. emissions of

households) per person (per capita) living in an area. Compute the growth rate of per

capita emissions between 2010 and 2015, for each area (Recall that regid is the area

identifier). Draw a histogram of this growth rate across areas. How much did

emissions grow in the area with the smallest per capita emissions growth between

2010 and 2015?

Explain which STATA commands you used to answer these questions and try to be as

efficient as possible (i.e. the less commands you use the better)

[6 points]

(b) The variable pct_leave contains the percentage of votes for leaving the EU in the

referendum on June 23, 2016. What is the relationship between per capita CO2

growth between 2010 and 2015 and support for Brexit? (hint: consider CO2 growth as

dependent variable)

[6 points]

(c) Consider two areas (Brexton and Remainster) whose per capita pollution is equal to 2

in 2010. Support for leave was 60% in Brexton and 40% in Remainster. What do you

expect the percentage difference in per capita emissions between Brexton and

Remainster to be in 2015?

[8 points]

Question2 (20 points)

Continue using the dataset brlongmerge.dta from Question 1.

The variable urate is the local area unemployment rate in percent. domCO2Opop is the

amount of CO2 emissions in tons) per person living in an area. The variable regid is a unique

regional identifier.

(a) Regress domestic per capita emissions (domCO2Opop) on the unemployment rate

(urate). Report the regression result. Can you motivate the causal relationship implied

by the regression? [4 points]

(b) Assuming that two areas differ in their unemployment rate by 10 percentage points.

How big a gap in per capita emissions do you expect between the two areas based on

the regression above? [4 points]

(c) A colleague suggests using year controls in the regression from a). Implement this

and report the results? . Explain if and why this might provide a better estimate of the

effect of unemployment on CO2 emissions. [4 points]

(d) Explain why the regression in c) might or might not contain include a dummy

representing the year 2015 as regressor. If you have not already done so, Provide an

interpretation of the coefficient associated with the 2015 time dummy variable in the

regression above. [4 points]

(e) Can you propose and motivate a further modification to the regression approach in c)

that does not involve adding any further variables to the regression but will make a

causal interpretation of the unemployment coefficient more plausible? Can you

implement this? Report and discuss the results. [4 points]

Question3 (20 points)

The table above is from a study on the effect of government support for disadvantaged areas

(wards) in the UK on employment. The policy consists of subsidies for firms investing in

those areas. The policy is variable the maximum amount of support firms can get in a

particular area, the maximum investment subsidy or NGE.1 This can go from 0 to 0.35; i.e. a

firm investing £100K will get £35K from the government. The columns of the table refer to

different regressions related to this policy. The regressions are reported in terms of

differences between 2003 and 1997; e.g. the first regression in column 1 reports a simple

OLS regression of the form

ln ???&''() – ln ???,--.) = ? ???&''() ? ???,--.) + ?&''() ? ?,--.) 1 NGE stands for net grant equivalent.

where ? indexes an area.

(a) Based on the OLS regression, what would be the growth of employment in an area

where between 1997 and 2003 NGE increased from 0 to 0.1 (i.e. 10% of investment

firms undertake will be paid for by government)? [5 points]

(b) What concerns – if any – would you have with the result above? [5 points]

(c) The researchers claim to have an instrumental variable for the policy variable,

referred to as “Policy Rule Instrument” above. What are the requirements of an

instrumental variable? What can you say (and what not) about the quality of the

researchers’ instrumental variable on the basis of the regression table above? [5

points]

(d) Assuming the instrumental variable strategy of the researchers is sound, what do you

consider to be the most correct estimate of the impact of 10% increase on

employment growth based on the table above? [5 points]

Question4 [20 points]

The following is an OLS regression result from a study on civil conflicts in African countries:

The dependent variable here is a dummy variable that is equal to 1 if a conflict occurs in a

given country that leads to more than 25 deaths per year. The explanatory variables are the

annual growth of rainfall in the current (i.e. t) and in the previous year (i.e. t-1).

a) What does the regression suggest is the impact of an increase in rainfall growth of

1 percentage point on the likelihood of civil conflict? [4 points]

b) What could be the reason for the observed relationship? [4 points]

c) Do you think this is a good estimate of the causal impact of rainfall growth on

conflict. Explain your answer. [4 points]

d) The main interest of the study is the effect of economic growth on conflict. One

approach could be to simply regress conflict on economic growth; i.e. run a

regression of the form:

?????????@ = ?,???????? ???????@ + ?&???????? ???????@D, + ??@

Discuss the merits of this approach. [4 points]

e) The authors of the study propose to use growth in rainfall as an instrument for

economic growth. Do you think this is a sensible approach? Explain your answer.

[4 points]

Question5[20points]

Consider the dataset proddata_clean.dta which contains data for a selection of European

regions (at the so called nuts2 level; for more details see

http://ec.europa.eu/eurostat/web/nuts) of regional (gross) value added (gva), capital stock and

employment. For each region we have data from 2000 to 2012. Productivity of regions or

countries is often measured by looking at the value added per employed person (gva/emp).

However, this does not take into account how much capital is being used in the production

process. Economists therefore prefer to use production functions where gva is modelled as a

function of employment (L) and capital (K)

? = ??HI?HK

In terms of assessing productivity economists will subsequently look at what they call total

factor productivity (TFP) represented by the letter A in the formula above.

Note: regions are uniquely identified by a numeric identifier (regid) or by the string identifier

nuts2. The variable country reports which country a region belongs to using 2 letter country

More details on those under: http://ec.europa.eu/eurostat/statistics?explained/index.php/Glossary:Country_codes

(a) Suggest and discuss an approach to estimate the function above. Try to deal as much

as you can with potential endogeneity issues. Implement this approach and report the

result estimates for the parameters in the function.

[4 points]

(b) Would you say it is reasonable to maintain the hypothesis that the production function

has constant returns to scale; i.e. ?M + ?N = 1? Explain your answer.

[4 points]

(c) Based on your estimate, compute for all regions and years the level of TFP (i.e. A) in

the function above. Explain what you are doing.

[4 points]

(d) For 2012 only examine which country had the lowest (and highest) average

employment weighted GVA per capita across regions? Which country had the lowest

(and highest) average employment weighted TFP level? Explain your calculations.

[4 points]

(e) For 2012 only, which country had the highest ratio in the GDP per capita between the

most and the least productive region; i.e. in which country was the biggest inequality

between regions in terms of GDP per capita. Compute the same result for TFP.

[4 points]


版权所有:留学生程序网 2020 All Rights Reserved 联系方式:QQ:99515681 电子信箱:99515681@qq.com
免责声明:本站部分内容从网络整理而来,只供参考!如有版权问题可联系本站删除。