Dear all,
I need to model a gdp per capita steady-state for a group of economies but I'm not sure about how to do that. I thougt in take the average of the filtered series (using HP filter, for example). Is it a good idea? Is there references to sustent the use of this? Do you know other approaches?
Thanks a lot!
Dear Prof. Penna,
If you are trying to test the convergence of country GDPs percapita my research can offer a tangential answer in that if you look at my published coauthored paper on mathematical statistical pricing and also published papers on "Quantitative" Pareto Optimality of Stock Markets on my RG page you will see that at the disaggregated microstructure level stock price dynamics are affected by growth of percapita income (proportionate) and growth of percapita interest rates in orthogonal ways over consecutive planning horizons with verifiable incomplete asset market structures forming the instrumental econophysics variables for convergence through open investments, and hence it is questionable as to whether convergence actually can take place in the GDP space of economies at least till these structural econophysics instrumental variables have been physically developed and people are educated appropriately about rational investment plans. Quantisation does not only solve stock market mediated convergence problems. Hence in terms of Macroeconomic Policy school the evidence suggests conclusively in favour of the structuralist school. SKM QC
Dear Christiano,
there exist different macroeconomic models of growth. In some of them (endogenous growth models, see the book by R.Barro & X.Sala-i-Martin) growth can be unlimited and there is no steady state. Another question is asked: whether there is convergence across countries in growth rate of GDP.
If you are talking about GDP per capita, you probably assume population growth that may differ across countries. Most of macro models deal with constant population. Population growth is usually dealt by demographers, who also include different cohorts and this makes math rather complicated.
But you are probably doing empirical research. Then you first have to ask what is your baseline theoretical model that you want to test.
Dear Christiano, before starting to define specific mathematical and statistical aspects of the steady state, and independently by the specific objective of your analysis (per-capita GDP), you must consider some general questions about steady state.
First. In economics, steady state is not forced to the zero growth rate. Steady state need not the absence of growth. The general and common character of steady state is that of a system in which the relevant variables grow constantly at the same and constant rate; obviously this rate can be equal to zero, but it not to be perforce zero.
Second. If the growth rate is the same, and if it is constant:
- the “distances” between economic variables are invariant: the ratio of two variables remains constant
- nonetheless, the level of the economic variables can increase over time, even if the system is in steady state
Third. From another point of view, it can be argued that in the steady state all the variables involved tend to the respective potential level. This fact not imply that the growth rate is perforce zero: in fact, the potential level of a variable can change over time, with a rate commonly named "natural".
Fourth. The steady state needs the stability of the system, but the general stability and not the local one.
Fifth. They exist some “particular” definitions of steady state; one of them is that of Solow R and Swan T. They define the steady state as the situation in which investments equal to their depreciation: the stock of capital don't grows anymore. It is interesting to note that this definition imply absence of economic growth, but the growth rate of investments is not zero. There is another "particular" interpretation of the steady state, that in my opinion is interesting; I suggest you the following contribution:
Daly H. (2009) - From a failed growth economy to a steady-state economy, in Encyclopedia of Earth, eds Cutler, Cleveland
http://www.eoearth.org/view/article/51cbedd27896bb431f6942d8/
Coming to the operative question, that is define a steady state model, if you intend to identify it in the "real world", I think it is impossible: in this moment, it doesn't exist an economic system in steady state. Pay attention to another point. You suggest the use of the HP filter: this method is useful to distinguish the trend and the cyclical component of time series, but the trend is not automatically the potential level. It is more easy to build a theoretical steady state model: you can choose a model of economic growth characterized by the general stability, and impose to it the same constant growth rate for the variables of the system.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS