Salaries Grow Before Elections, Don't They?

One thing public opinion polls cannot fully show are the main reasons why people vote for certain parties or why they are undecided. In this article we will roughly present the main elements of the political economy of democracy, that is the inter-dependence of economic and political cycles in democratic countries, and thus we will also point to its relevance for our situation. As far as the below signed knows, this will be the first contribution to domestic writings on this issue. In market economies with a stable democratic electoral procedure it has been noticed that there is an inter-dependence between the government's conduct before the elections and the voters preferences on the one hand, and economic indicators on the other. When coming to power, a government increases unemployment in order to initiate a drop in inflation and then, somewhere at the middle of its term of office, it conducts an economic policy which leads to a drop in unemployment prior to the next elections. In this way it brings itself in the most favorable position regarding the voters' preferences, leaving it up to the future government to fight inflation which must follow after a drop in unemployment. As regards the conduct of voters, especially the undecided ones, it is assumed that they suffer from myopia, that is that they take into account the rates of inflation and unemployment at that very moment, without giving much though to their future values. Research conducted in countries like the United States, Great Britain and Germany have shown that the popularity of a government (usually determined by public opinion polls) is very much linked to the mentioned economic variables. One can ask how relevant the mentioned models of public choice are in the case of Yugoslavia. In these regions economic topics were pushed aside long ago. This is quite understandable from the standpoint of the authorities, due to the catastrophic results of the economic policy, but it is surprising that the opposition is ignoring the field of the economy where it can "beat" the regime with a wide range of arguments. We consider that if one takes a look underneath the heavy layers of national rhetoric which is now being used by both the ruling party and the opposition, one can find concrete economic interests which certain voters do pay attention to when making their choice. Precisely for this reason, we consider it necessary to set out certain empirical findings to show why we have chosen to discuss the political economy of democracy. In the border zone: The main methodological limitation of a more serious examination of the inter-dependence of the Government's policy pending the elections, that is the preferences of the voters, and economic cycles here, is the fact that these will be only the second (first?) free elections. By this we want to say that the available time series for our empirical research are relatively short. However, as the first step in an in-depth examination of the mentioned phenomena of political cycles, we can use a simple graph showing two key economic indicators which determine the preferences of voters in normal market economies. The first one shows inflation trends (measured according to the growth of retail prices), and the other one presents the trends of overall salaries in real terms per employee in the Federal Republic of Yugoslavia over the past three years, with a forecast of our own for the period until February next year. We have opted to follow salaries in real terms since, due to the institutionally different treatment of the labour force, the indicators of employment, that is unemployment, are less interesting because of their relatively small correlation with inflation trends, which is not the case in market economies. Another reason to leave out the facts about unemployment trends is their relatively late registration compared to the indicators pointing to inflation and salaries. However, we can conclude that the Government's decision that no one can lose a job while the sanctions are in force, will slow down the growth of unemployment. Even though pictures speak for themselves, it is necessary to add a few comments and figures which will further explain the presented trends.

INFLATION TRENDS

Percentage

1. elections

2. elections (?)

Source of information: the Federal Statistics Bureau

Achieved Forecast

In the last quarter of 1990, the average inflation rate accounted for 4.88 percent, only for an inflation rate of 7.74 percent per month to be registered right after the December elections, in the first quarter of 1991. Since the break-out of the war in Croatia in August 1991, we have been witnessing that the economy is once again entering the zone of hyper-inflation. Prices practically grew constantly for a year, until July 1992 when it started becoming clear that elections would be held in the course of the year. Since then, we have an economically unexplainable slowing down of inflation, of the kind we had in the first half of 1990, at the time when Ante Markovic's program functioned but within which the drop in inflation was made possible by coordinated measures of the economic policy. Then one paradox came after another: in an economy against which the world community has imposed a full economic blockade, with foreign trade flows having been suspended, in which production is drastically dropping and the number of seemingly employed is growing, prices keep dropping month after month! In the last three months of this year the average inflation rate accounted for 49 percent which is much less than the rate registered in the first half-year period (when the monthly average was 64 percent). According to trends before and after elections, we can conclude that, after the 1992 December elections, the prices will grow by more than half of their three-month average (growth rate of around 60 percent, with a tendency to grow further). Thus we would once again exceed the limit of 50 percent a month, which American economist Phil Cagan has taken as the border zone of hyper-inflation. It is interesting, in this context, to point to the ruling party's election slogan in 1990: "There is no uncertainty with us". It has been proved that this slogan is in accordance with the facts we are presenting, but only in regard to the period just before the elections. Namely, the statistical indicator which serves as a measure of uncertainty suggests that the growth of prices and salaries is the least variable precisely at the time of elections, and that after the elections the degree of uncertainty grows in the sense of the direction and the intensity of the growth of, that is drop in, prices and salaries. Thus a feeling of stability with the present authorities is suggested to the voters, precisely before the elections. The unexpected slowing down of inflation in the past three months can most probably be ascribed to non-economic factors. With this, the government counts on the mentioned shortsightedness of the voters who take into account only the last registered inflation and unemployment rates. When we speak about the unexplainable drop in the inflation rate, we have in mind precisely what Ljubomir Madzar, federal minister of the economy, also took into account when, at his first press conference he set out the hypothesis that the November inflation rate of 33 percent is "partly also a statistical phenomenon" ("Sunday's Borba", December 6th 1992). We would also like to express doubt in the methodological correctness of the procedure for determining the indicators of inflation. At the same time, we wish to believe that the "shaping up" of statistical data is left behind in our socialist past. Learning from experience: The trends of salaries in real terms illustrate to an even greater extent our claim that the Government has learned the first lessons in the political economy of democracy. Prior to the December elections in 1990, an average monthly growth rate of salaries of around 4.2 percent was registered, which is a fantastic result considering the fact that, in the first half-year period, salaries dropped in real terms at a rate of around 4.4 percent a month. At the time the current authorities also dictated the time of the holding of elections, assessing that it is precisely in December that they will be as close as possible to the voters' preferences. The results of the elections unequivocally confirmed this. The story is also being repeated with the 1992 December elections. After a catastrophic drop in salaries in real terms in the first six months of 1992, when the average monthly drop accounted for almost 11 percent, from July (when it became almost certain that early elections will be held) until October (the last available information) the average monthly drop in salaries in real terms accounted for only 0.8 percent. The Government is creating an impression with the voters that it is protecting their living standard despite the blockade, the drop in production, the growth of unemployment and other economic indicators with a negative trend. Bearing in mind the experience from the December 1990 elections, we assess that after the elections, until February next year, the drop in salaries in real terms will accelerate as presented on the graph.

TRENDS OF SALARIES IN REAL TERMS PER EMPLOYEE - TOTAL

Indexes (1991=100)

1. elections

2. elections (?)

Source of information: The Federal Statistics Bureau

Achieved Forecast

And what to say at the end? According to the mentioned theoretical model, the voters' expectations are formed in an adaptive way. This means that voters learn from their previous experience. The two presented graphs clearly show what the Yugoslav voters can learn from the experience from the December elections. If the current regime stays in power, already at the beginning of the year there will be an enormous growth of prices and a drastic drop in salaries in real terms. For this reason, the key question being asked is: did our voters learn this already after the first elections, or will they, like bad pupils, have to repeat the lesson?