Wednesday 21 November 2007

Tell me, what is real?

Chapter nine of the WinEcon program, the first of the macroeconomic sections, turns immediately to some of the basic issues and indicators—interest rates, GDP growth, inflation and the balance of payments—with which macroeconomics deals, as well as some of the tools used to manipulate, describe and analyse them. This is used as a way of introducing some important definitions and distinctions concretely.

Kinds of variables
The first such distinction is between variables that show a pattern of increase or decrease over time (trended variables) and those that show no consistent long-term pattern (non-trended variables). The former include GDP and its components, inflation and the money supply; the latter include ratios such as unemployment rates, interest rates and exchange rates. A second basic distinction is between nominal variables, which measure changes in value, and real variables, which measure changes in volume or physical quantities. Ratios constitute a third category of variable.

Calculating real interest rates
This distinction between nominal and real is investigated in relation to interest rates, which indicate the scale of return on lending, or the cost of borrowing. A nominal interest rate is a ratio of asset yield to asset value (the asset could be a loan, for example), whereas the real interest rate—which takes into account movements in the price of goods and services—is a measure of the extra quantity of goods and services that can be bought from the loan plus the return. An approximate measure for calculating real interest rates is therefore the nominal interest rate minus the rate of inflation; the formula for a more accurate calculation is as follows:
Real interest rate (RR) = (1 + nominal interest rate/1 + inflation rate) -1
Therefore, if the annual nominal interest rate is 10% and the annual inflation rate is 6%, the crude measure gives a real interest rate of:

RR= 10 - 6 = 4%
Using the more accurate measure, the calculation is as follows:

RR = (1 + 0.1/ 1 + 0.06) -1 RR = 0.0377358—ie 3.77%
That is, rather than getting 10% more goods and services for the 10% return on your loan, with an inflation rate of 6%, you only get 3.77% more goods and services.

High inflation rates can turn real interest rates negative, with potentially profound economic consequences—helping to explain fluctuation in investment across time, for example. The pattern of interest rate developments in a selection of developed Western countries over three decades shows, broadly, that real interest rates were:
  • low but positive in the 1960s;
  • negligible to negative in the 1970s; but
  • relatively highly positive in the 1980
Detrending GDP
Next, the program looks at three ways for analysing the patterns of economic growth, introducing the idea of detrending—a method by which any cyclical changes in trended variables can be separated out from the trend itself (this is useful when looking at short-term fluctuations in aggregate data).

Deviation of GDP from the linear trend. One method for detrending is to fit to observations of the level of GDP a line that minimises the sum of the squared deviations. The pattern of deviations from this line, when graphed, brings out the periods in which growth is above or below trend. (The formula for calculating the percentage deviation from trend level GDP is as follows: [real GDP - trend GDP/ trend GDP] x 100.) Looking in this way at the components of GDP by expenditure for the UK for 1955-2004, we can see that:
  • household consumption grew at a rate below trend between 1975 and the late 1980s, but above trend since the late 1990s;
  • government consumption was below trend for most of the 1990s, but above trend since the coming to power of the Labour Party in 1997;
  • the growth of exports has been above trend throughout the 1990s; and
  • the pattern of investment growth has been more erratic.
Deviation of natural log of GDP from the log-linear trend. If a variable grows at a constant rate, however, a second method can be used to get an accurate picture of patterns of change separate from the broad trend, which is to log the trend to produce an analysable linear graph, and then calculate the deviation of the logged GDP data from it. (We calculate the percentage deviation from a log-linear trend as follows: [natural log of GDP - log of trend] x 100.)

Period GDP growth rates. A third, more widely used, approach for separating out patterns of change in trended variables from the trend itself is to use growth rates of GDP in each period, with periods of high positive growth indicating a boom, and periods of contraction (typically, three quarters in succession) indicating a recession.

Constructing a price index
Inflation is the rate of change of the average price level (whereas hyperinflation is defined as persistently high rates of inflation, typically stemming from the inability of the authorities to tax and borrow amid generalised societal breakdown, often as a result of war, civil war or revolution). To get a realistic picture of inflationary patterns, it is important that the weighting of the goods and services in the price index reflects their relative importance in current patterns of spending, since these change over time. The first method, a base-weighted index, asks the question: How much do you have to spend, measured in the prices prevailing in each subsequent period, to buy the same quantity of goods as in the original, base period? A second method, a current-weighted index, asks: What would spending be if prices had stayed the same? The important GDP deflator is a ratio of the sum of the current values of all goods and services that make up GDP to the sum of the value of the same output in constant (base-period) prices, and it is used to work out real rates of economic growth.

The impact of devaluation on the balance of payments
The balance-of-payments records an economy’s transactions with the outside world, and may be used, among other things, to produce a picture of the geographical distribution of a country’s most important markets and suppliers. The balance-of-payments accounting framework may also be used to analyse the likely impact of using a devaluation of a fixed-exchange rate to address persistent external deficits—which should make exports cheaper and imports more expensive, stimulating and discouraging them, respectively, in order to close the external gap. Looking at the empirical results of just such a devaluation in the UK in the late 60s, we see that the external position temporarily worsened before it improved (the so-called J-curve), because both foreign and domestic demand takes time to adjust. In the short run, export quantities and prices in local-currency terms are unaltered; import quantities too are fixed, but local-currency prices for foreign goods and services rise straight away, so that the external imbalance is initially exacerbated. In the long run, however, export quantities increase and those of imports fall, as expected, reducing the external deficit.

Co-movements
A final section touches on the interesting and important topic of the interaction between macroeconomic variables, both within the domestic economy and between national economies. Plotting a scatter point diagram of GDP growth against a number of other domestic economic variables, the degree of correlation between them is indicated by the closeness of the co-ordinates to the best-fit line through the scatter points.
  • A best-fit line with a positive slope indicates a pro-cyclical relationship—that is, one in which both variables move in the same direction. Those variables with the strongest pro-cyclical relationship with economic growth empirically include consumption, investment and imports.
  • A negative slope indicates a countercyclical relationship, which means that a rise or fall in GDP growth is associated with a fall or rise, respectively, in the other variable. Empirically, changes in the rate of unemployment and inflation exhibit just such a relationship.
  • When the best-fit line is horizontal, the relationship is described as acyclical, and no systematic relationship is discerned.
Correlations are scored between +1 and -1; the closer the correlation is to +1, the closer the scatter points to the positively sloping best-fit line, with a score of +1 indicating a perfectly positive correlation, in which all of the scatter points coincide with the best-fit line.

However, the linked movements of one macroeconomic variable and another are not always simultaneous. Sometimes a change in a variable occurs before a change in economic growth, and leads it; sometimes the change happens after a change in GDP, lagging it. Making an allowance for possible leads and lags in this way helps to untangle evidence regarding the difficult question of cause and effect.

GDP growth and interest rates. The relationship between economic growth and the Treasury-bill rate shows a strongish countercyclical lead movement, suggesting that an increase in the interest rate is associated with reduced output growth in the later period. The contemporaneous relationship is weakly pro-cyclical, but this strengthens when the variables are lagged, yielding some evidence that increased output growth pushes up interest rates, although with a delay.

GDP growth and changes in the unemployment rate. There is strongish countercyclical relationship when the variables are contemporaneous, which strengthens when the variables are lagged—that is, the rate of unemployment falls with rising economic growth, and this effect becomes more powerful over time.

GDP growth and inflation. An empirical investigation between output and inflation in the UK in 1951-93 reveals the interaction of the two to be slightly more complex than expected—specifically, there is a positive short-run connection between output growth and lagged inflation, but with periodic shifts, so that increases in output growth in the sub-period 1975-82 are associated with much larger increases in lagged inflation than for either 1951-74 or 1983-92.

Money supply and inflation. The strongest positive correlation between the rate of growth of the money supply is seen when plotted against data for the rate of inflation lagged for two years.

GDP growth and growth in other economies. Here, the question is: To what extent are booms and recessions transmitted? This depends on the degree of openness of a national economy to world trade and who its main trading partners are, which is often linked to geographical proximity. Thus, the GDP growth of the UK is most strongly positively correlated to economic growth developments in the US, Japan and France; for the US, GDP growth is most strongly linked to economic growth in Canada; for France, Italy; for Japan, Germany.

Monday 12 November 2007

The Triumph of the West

After some time away from blogging—mostly trying to get to grips quickly with the politics and economics of the Kyrgyz Republic—I intend to get back on track with my studies of maths and economics more broadly, starting with a critical review-stroke-summary of How Capitalism Was Built, a new book by Anders Åslund in which he outlines the main sequence of events of the transition process in eastern Europe and the former Soviet Union, as well as the theoretical and policy debates linked to them.