A carregar...
Publicação
Evaluating tests for convergence of economic series using Monte Carlo methods with an application to real GDP's per head.
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 3.29 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
The convergence concept can be found in different fields of the Economics discipline.
Convergence of incomes or ofGDPs per head is an important issue in growth theory. Some
models of international trade imply the convergence of factor prices across countries. The
European Union treaty includes some convergence rules on interest, exchange and inflation
rates, and on budget deficits, for countries to enter a .monetary union. This thesis starts by
reviewing this literature.
These developments are accompanied by a number of empirical studies concerning the issue
of measuring and testing for convergence. Proposed methods and provided results are in
apparent contradiction. They are critically surveyed in Chapter 2.
The aforementioned disparate results constitute one of the main motivations for the systematic
evaluation of the different methods. This is done resorting to simulation techniques using
artificial data. Different techniques are assessed considering a number of different patterns of
convergence. Chapters 3 to 7 include several experiments considering unconditional and
conditional convergence, convergence clubs, limited and time-varying convergence as the
true data generation process. Their results allow a better understanding of previous empirical
studies and of the comparative advantages and weaknesses of different tests. In general terms,
cross-sectional methods are not very reliable in the presence of cross-sectional heterogeneity.
Time series methods do not share this disadvantage. A Kalman filter method is more robust
to a time-varying speed of convergence when compared to cointegration techniques.
The thesis includes an empirical investigation on the convergence of GDPs per head across
16 industrialised countries using annual data from 1890 to 1989 (Chapter 8). The main
conclusion is that countries tended to converge conditionally towards the US level, specially
after the Second World War, at different speeds and to steady-state levels that are different
from the pre-war values.
Descrição
Doutoramento em Economia.
Palavras-chave
Contexto Educativo
Citação
St. Aubyn, Miguel Pedro Brito (1995). " Evaluating tests for convergence of economic series using Monte Carlo methods with an application to real GDP's per head. University of London. London Business School.