ECONOMIC DEVELOPMENT. K. Harris: Candidate M.A. Applied Economic Analysis

Transcription

1 ECONOMIC DEVELOPMENT K. Harris: Candidate M.A. Applied Economic Analysis

2 Introduction PRESENTATION Section 1 Data Driven Section II Replication Section III - Extension Section IV Summary Section V Food for Thought

3 INTRODUCTION THE WORLD AS WE KNOW IT

4 WHERE DO WE BEGIN? $ Billions in aid to Haiti to rebuild after earthquake. Most of the $$$ remain unspent Haiti s GDP per capita $ 819 USD

5 WHERE DO WE BEGIN? $ Billions in aid for many years including after the earthquake What has changed? Where is the money? Nepal s GDP per capita $ 694 USD

6 STUDY HISTORY? Look for major event over time The Great Depression Herbert Hoover vs Franklin D. Roosevelt

7 Civilian Conservation Corps Civil Works Administration Federal Housing Authority Federal Security Agency Home Owner s Loan Corporation National Recovery Act Public Works Administration Social Security Act Tennessee Valley Authority Works Progress Administration PUBLIC POLICY- THE TRANSFORMATION OF AN ECONOMY Nothing to Fear by Adam Cohen, FDR s New Deal (the creation of the Modern America?)

8 WHERE ARE WE GOING?

9 SECTION 1- DATA DRIVEN

10 METRICS World Bank s World Development Indicators GDP per capita Income distribution Literacy and education Access to healthcare Social Security and pensions Modern transportation

11 DATA Penn World Tables (PWT Mark 5) An expanded set of international comparisons. The PWT displays a set of national accounts economic time series covering many countries. Its unique feature is that its expenditure entries are determined in a common set of prices in a common currency so that real international quantity comparisons can be made both between countries and over time. In addition, it presents data on relative prices within and between countries and demographic data and capital stock estimates as well. Summers and Heston

12 DATA PWT 5.6 Data Dependent variable Real GDP per capita (chain): RGDPCH A chain index obtained by first applying the component growth rates between each pair of consecutive yare, t-1 and T, to the current price component shares in year t-1 to obtain the per capita domestic currency (DA) growth rate for each year. The DA growth rate for each year, t, is then applied backwards and forwards and summed to the constant price net foreign balance to obtain the Chain GDP series. Predictor variable The ratio of fixed capital formation to GDP: INV fixed capital formation

13 MODEL Identify the structure of the data first to apply the appropriate model Panel data cross-section dimension and time dimension y it = β 0 + x it + v it Fixed Effects (FE) Model If only interested in time-variant factors y it = β 0 + x it + a i + u it Random Effects (RE) Model Variation across entities is assumed to be random and uncorrelated with independent variables y it = α + βx it + u it + ε it

14 GRANGER PREDICTIVE CAUSALITY Concept of causality in time series analysis. More difficult in panel data A variable X Granger-causes Y if Y can be better predicted using the histories of both X and Y rather than just the history of Y RGDPCH t = RGDPCH t RGDPCH t-2 RGDPCH t = RGDPCH t RGDPCH t ci t-1 Interpret investment to be Granger-causing growth when a prediction of growth on the basis of its past history can be improved by taking into account the previous period s investment

15 GRANGER PREDICTIVE CAUSALITY

16 GRANGER PREDICTIVE CAUSALITY Ci t = ci t ci t-2 Ci t = ci t ci t RGDPCH t-1 Interpret growth to be Granger-causing investment when a prediction of investment on the basis of its past history can be improved by taking into account the previous period s growth

17 GRANGER PREDICTIVE CAUSALITY

18 CONCLUSION Using simple causality tests there does not seem to suggest capital formation induces growth nor does growth induce capital formation

19 SECTION III - REPLICATION

20 REPLICATION Is Fixed Investment The Key To Economic Growth? By Blomstrom, Lipsey, Zejan Based upon a NBER # 4132 What Explains Developing Country Growth? PWT 5.6 data used 5 year periods to calculate RGDPC as ration, t 5 / t 1 used 5 year average for investment

21 REPLICATION

22 REPLICATION

23 REPLICATION Fixed Effects Model Covariance model Within estimator Individual dummy variables model least squares dummy variable model Random Effects Model Random intercept model Partial Pooling model

24 SECTION III - EXTENSION

25 EXTENSION

26 SUMMARY Results did not match the authors results. I used Fixed Effects (FE) Model FE was chosen after testing various models with Hausman RGDPCH chain index drgdpch 5 year period, t 5 / t 1 yravgci 5 year average: (Σ ci t )/5 Modeled Fixed Effects (FE) Model Random Effects (RE) Model Least square dummy variable (LSDV) Model

27 FOOD FOR THOUGHT At the time, I had little idea that so many people had very fixed ideas about causation, but they did agree that my definition was not true causation in their eyes, it was only Granger causation. I would ask for a definition of true causation, but no one would reply. However, my definition was pragmatic and any applied researcher with two or more time series could apply it, so I got plenty of citations. Of course, many ridiculous papers appeared. Nobel Lecture, December 8, 2003 by Clive W.J. Granger

28 APPENDIX

29 BREUSCH-PAGAN LM

30 HAUSMAN FE OR RE MODEL

31 TESTING FOR TIME-FIXED EFFECTS

32 TEST FOR CROSS-SECTIONAL DEP

33 RANDOM EFFECTS MODEL

34 FIXED EFFECTS MODEL

35 FE WITH DUMMY VARIABLES

36 REFERENCES