"First impressions are lasting impressions." The first time I met Professor Buchanan was in 1964 at the Brookings Institution. Fresh out of graduate school, a Musgravian by training, I was recruited to write (with David J. Ott) a book on federal budget policy. At the time, I believed in the "Welfare State", that the role of the government is far beyond the "minimal state". In the Brookings dining room, I found myself sharing a lunch table with Professor Buchanan and a few members of the government study program. I don’t remember how I got to sit there or who else was there, but one thing I remember was the presence of Professor Buchanan. His views about the public economy, the rights and the responsibility of the individual in society, contrasted sharply with what I held as the "learned truth". It was not until later on, in the 1970s that the seeds he sowed came to bear fruit. I began to question what I had learned and chart a course for myself.

Over the years I have come to know and value the teachings and writings of Professor Buchanan. What prompted me to write the paper "The Seventh Vote", as a tribute to his 80^th birthday is another, more recent encounter, again around a dinner table. At that meeting, I discovered that we share a mutual admiration for a British author, Nevil Shute. During the conversation, I found out that the quality that has attracted me most to Shute’s novels also endeared him to Professor Buchanan. In all his novels, Shute celebrates the "decency of the ordinary man".

I have selected one of Shute’s novels "In the Wet" (NY: William Marlow & Co. 1953) as it is Professor Buchanan’s favorite as the basis for this paper. The paper is co-authored with my former student Ron Shadbegian, Associate Professor of Economics, Umass Dartmouth, MA. The scene of Shute’s novel is Lansborough, in the Gulf Country, Australia. The time frame is 30 years forward.

‘Stevie’, I said, ‘I am Roger Hargreves. You know me; I am the person from Landsborough. Remember me?’

‘I am bad…I have been bad three days.’ ‘We’ll get you into Landsborough tomorrow, and then the ambulance will come and fly you to Cloncurry.’

‘I have flown further than that,’ he mattered. ‘Up’n down, up’n down, all across the world, carrying the Queen, Ottawa, Keeling Cocos, Nanyuki, Ratmalan. I know all them places. I got the seventh vote – did you know that, Cobber?’ ‘Did you know I got the seventh vote?’

Between dream and reality, a man in a coma recounts a memory of a life lived or to be lived. He is in England in the RAAF. His dinner host Captain Osborne asks: ‘There’s one thing about Australia I wish you would tell me. How does your multiple vote work? Have you got more than one vote yourself?’

The pilot nodded. ‘I am a three-vote man.’ ‘What do you get your three votes for?’ The captain asked.

‘Basic, education, and foreign travel.’

‘The basic vote – that’s what everyone gets, isn’t it?’

‘That’s right.’ The pilot said. ‘Everybody gets that at the age of twenty-one.’

‘And education?’

‘That’s for higher education, there is a whole list of other things, like being a solicitor or a doctor. Officers get it when they are commissioned. That is how I got mine.’

‘And foreign travel?’

‘That’s for earning your living outside Australia for two years.’

‘You can get more votes than three can’t you? Is it seven?’

‘The seventh is hardly ever given.’ ‘Only the Queen can give that.’ ‘The others are straight forward.’

‘You can get a vote if you raise two children to the age of fourteen without getting a divorce. That’s the family vote.’

‘The fifth is the achievement vote.’ ‘You get an extra vote if your personal exertion income, what you call earned income – in the year before the election is five thousand a year.’ ‘It is supposed to cater for the man who’s got no education and has never been out of Australia and quarreled with his wife but built up a big business. They reckon that he ought to have more say in the affairs of the country than his junior typist.’

‘And the sixth?’

‘That’s if you’re an official in the church,...you don’t have to be a minister, what it boils down to is that you can get an extra vote if you’re doing a real job for the church.’

Adapted from "In the Wet",1^st Edition, pp. 97-99.

One man, one vote is the premise upon which rights are allocated in a democratic society. Under majoritarian rule, the political outcome is said to be determined by the median voter. One implication of the median voter model is that representative democracies tend to be dominated by two political parties. When there are more than two, a coalition government is often formed between parties closely spaced on the preference side of voters. The two models give rise to different fiscal outcomes. Using the social costs of adjustment as the criteria, Spolaore (1993) contrasts the behavior of three types of governments: a social planner, a single party government, and coalition governments. Following an exogenous shock, a social planner acts immediately to minimize the cost of stabilization. A single party government comes second best in that it too avoids the waiting costs, but it deviates from the social optimum as it attempts to shift the burden of adjustments to the opposition party. A coalition government fares worse than either the social planner or the single party government since not only does it distribute the fiscal burden sub-optimally, but it does so with delay.

If one were to accept the proposition that a single party government is preferable to a coalition type government, one might ponder the possibility of a voting system, that would give rise to a single party government. One such system is the multi-vote system described by Shute’s "In The Wet".

‘I remember when the multiple vote started… It was when I was in Townsville, in 1963. They brought it in for West Australia.’

‘Why did West Australia start it?’ asked Rosemary. ‘Why not New South Wales, or Queensland?’

‘I don’t know,’ … ‘Labour was very much against it.’

‘West Australia was always pretty Liberal,’ the pilot said. ‘People had been talking about multiple voting for a long time before that. I reckon it was easier to get it through in West Australia.’

‘How did it come to be taken up by the other States, if Labour was so much against it?’ asked Rosemary.

‘Aw, look,’ said the pilot. ‘West Australia was walking away with everything. We got a totally different sort of politician when we got the multiple vote. Before that, when it was one man one vote, the politicians were all tub-thumping nonentities and union bosses. Sensible people didn’t stand for Parliament, and if they stood, they didn’t get in.’ (p.101).

In this paper we put Shute’s voting proposal to empirical test. We ask two questions: Will the one man one vote outcome be negated? And secondly, which was the pivotal vote? Was it the second, the third, the fourth or the fifth?

In the next section of the paper, we discuss the data and methods. The following section presents the findings. We would like to note from the outset that this exercise was made to satisfy intellectual curiosity. We have selected the German federal election of 1976 rather than more recent ones so that no political inferences can be attached to our findings.

2. Data and Methods

The data used to explore the implications of a multi-voting system are derived from the 1976 German election panel studies.¹ The election studies consist of national area probability samples of voting age citizens of the Federal Republic of Germany (FRG) excluding West Berlin. Oral interviews were conducted in three waves: two preceding the election, and one immediately following the election.

The panel interview dates for the 1976 election study were: May/June, August/September and October/November. The sample size was 2076 for all three waves. The pre-election interviews elicit information on respondents’ assessments of their "present" and "future" economic situation as well as that of the country. In addition, information was gathered about respondents’ party affiliation, perceptions of party position on issues, their voting behavior in previous elections, as well as expected vote in the next election. In the third wave, respondents’ actual votes are recorded along with party identification. Socioeconomic characteristics of respondents are given in the personal data section of the three waves. Weights are given for each variable so that sample distribution of characteristics corresponds to that of the total population. Since the German electoral system differs from our own (US), a few words about the German system are in order.

The Federal German Republic (FRG) electoral system is based on two principles: proportional representation and relative majority of single member constituencies. The 496 (plus 22 for Berlin) seats in the Bundestag (the lower house of the parliament) are divided one half each between "constituency candidates" and "party lists". The voting system allocates to each individual two votes: the first vote is cast for the "candidates" in a constituency. The candidate gaining a simple relative majority is elected. In this way one half of the members of the Bundestag are elected. The second vote goes to party lists, each party having a separate list in each of the Lander (state). The second half of the members of the Bundestag are elected from these lists in proportion to the number of second votes gained by each party. Under this electoral system, political parties compete both for the constituency votes and for their party lists. Given that each voter has two votes, it is of interest to ascertain whether both votes are cast in favor of one party or that voters split their votes among candidates of different parties. Evidence suggests that vote splitting is prevalent in the German election.² However, the decisive vote, the vote that shapes the outcome, is the second/party vote (Zweitstime) as it is the party vote that determines the final percentage of seats in the Bundestag. The seats won in the first vote are deducted from the total allocated to the party on the basis of the proportionality representation principle.³

The major political parties are: the Social Democratic Party (SPD); the Christian Democratic Social Union (CDU/CSU)⁴, the Free Democratic Party (FDP) and the GREENS.⁵ The SPD is the oldest political party dating back to 1875. It has a large membership (1,022,000 in 1976 compared to 798,000 for CDU/CSU) consisting mainly of urban working class. The CDU was founded in 1945 by center-right and center-left politicians, who advocated a "Christian" approach to politics and social issues. The party endorses the concept of a free market economy, although it stresses the need to meet "market failure" in the area of social policy. The party is identified with the business community, with little support from the working class. The FDP seems to combine elements of both conservative authoritarian views with strong ties to business, with left-wing factions that espouse the principles of democracy and individual liberty. On economic issues, it stood to the "right" of the CDU/CSU, (its coalition partner from 1949 to 1956 and from 1961-1966). Since 1968 the party has moved to the left, forming a coalition with the SPD since 1969. The Green party is a newcomer to the political scene. It has been identified with the protest and environmental movements of the 1960s. The Greens party advocates the re-orientation of social and economic life to reconcile man and nature. The party made its first successful entrance into politics in 1983 by exceeding the exclusionary rule of 5%. In 1983, they received 5.6% of the vote and 8.3% in 1987.

In the next two tables (table 1 and table 2), we provide information about the voting behavior of the survey respondents of the 1976 German election panel study as well as the actual votes in 1976.

‘So you got three votes. How does that work out in practice, at an election?’

‘You get three voting papers given to you, and fill in all three, and put them in the box,’ the pilot said.

‘You’re on the register as having three votes?’

‘That’s right. You have to register again when you get an extra vote – produce some sort of certificate.’ (p.101)

To simulate the effect of multiple votes on the election outcome, the following procedure was followed. First, we have to identify those individuals who would be eligible to receive the extra vote(s). The German election panel survey of 1976 identifies four characteristics that give rise to extra votes. These are: education, profession, income, and service to the church. The other two characteristics that are listed by Shute are ‘foreign travel’ and ‘raising a family’, but these could not be ascertained from the data and as for the seventh vote, only the "Queen" can give it. The selection criteria as applied to the sample are given in table 3. Sample means and standard deviations for the characteristics are shown in table 4.

Once those eligible to receive an extra vote(s) have been identified, the next step was to record for each individual the number of extra votes he/she is eligible to receive. We then calculated the distribution of votes cast for the parties on the assumption that the ‘extra vote(s)’ is (are) cast for the same candidate chosen initially by the first vote and also for the same party list chosen by the second vote.

As the data does not allow us to "replicate" the actual process of determining a party’s representation in the Bundestag (no information is given in the sample on the proportionality representation), we have assigned seats to each party on the basis of the average of the two votes (candidate and party lists) cast for that party. The voting outcome under the new system is given in table 5.

From the data reported in the table, the election has been "overturned". CDU/CSU commands a majority. Therefore on the basis of votes cast, a single government replaces the previous coalition government which was made up of SPD and FDP.

An analysis of the direction of effect as well as the contribution of each of the extra votes to the outcome is given in tables 6, 7 and 8. Table 6 gives the sign and significance level, table 7, the logit coefficients and table 8 the partial derivatives of the logit estimates evaluated at the sample means.

	A.	FIRST	VOTE
Variable	Coeff	SPD b/s.e	Mean of x	Coeff	CDU/CSU b/s.e	Mean of x
COLLEGE	-0.165	-2.858*	0.058	0.062	1.154	0.058
RICH	-0.035	-1.303	0.244	0.022	0.830	0.244
PROF	-0.044	-1.272	0.132	0.069	2.00*	0.132
HOLY	-0.288	-6.448**	0.108	0.306	7.253**	0.108
	B.	SECOND	VOTE
COLLEGE	-0.130	-2.326*	0.241	0.057	1.083	0.059
RICH	-0.045	-1.704*	0.059	0.021	0.796	0.242
PROF	-0.021	-0.603	0.132	0.054	1.580	0.132
HOLY	-0.296	-6.572**	0.111	0.304	7.326	0.111

The logit results are about as one would expect. CDU/CSU was expected to gain votes under Shute’s multiple votes, especially the HOLY (service to church) vote. From the partial derivatives, we can allude to the effect of each of Shute’s votes on the probability of voting for a particular party. The two most significant votes are COLLEGE and HOLY. The results of the logit analysis show that the college vote reduces the probability of the first vote going to SPD by 16.5%, while the HOLY vote reduces this probability by 28.8% and increases the probability of voting for CDU/CDC by 30.7%. For the second vote, the church service or HOLY vote reduces the probability of voting SPD nearly as much as it increases it for the CDU/CSU, a reduction of 29.6% for SPD and a rise of 30.4% for CDU/CSU. COLLEGE and RICH both have negative effects on the probabilities of voting SPD, but little effect on voting for CDU/CSU.

Turning now to the second question: Which of the extra votes turned out to be the "pivotal" vote for the CDU/CSU win? We have tried several combinations with the following outcomes:

1. PROF, COLLEGE and RICH < 50% for CDU/CSU
2. PROF, HOLY = 50% for CDU/CSU
3. PROF, COLLEGE is slightly < 50% for CDU/CSU
4. PROF, RICH slightly < 50% for CDU/CSU
5. HOLY, PROF, COLLEGE = 50.3% for CDU/CSU
6. HOLY, PROF, RICH = 50.3% for CDU/CSU

Clearly then, the pivotal vote is the church (HOLY) vote. No matter how the extra votes are combined, CDU/CSU wins a majority only when that vote is included.

‘And after the multiple vote came in, was it different?’ Asked Rosemary.

‘My word,’ said the Australian. ‘We got some real men in charge. Did the Civil service catch a cold! Half of them went out on their ear within a year…’ (p.101).

Nevil Shute, "In The Wet", (p.101).