When seeking support for the League of Nations, American President Woodrow Wilson, declared that he wanted to make the world safe for democracy. Early theorists of democracy by contrast wanted to ensure that democracy could be made safe for the world. They were particularly apprehensive of the dangers of untrammelled leadership. Indeed, Jeremy Bentham became converted to the virtues of democratic government precisely because it was only in a democracy, so he believed, that leaders could be held to account. Tocqueville, like Mill, feared a tyranny of the majority, a majority in thrall to a leader who had secured control of the state. ‘As each member of the community is individually isolated and powerless,’ he wrote, ‘no one of the whole body can either defend himself or present a rallying point to others; nothing is strong in a democratic country except the state’. But ‘the people’ would nevertheless rouse themselves, ‘shake off their state of dependence long enough to select their master, and then relapse into it again’. Tocqueville predicted democratic despotism, a prediction that became appallingly true when Adolf Hitler became Chancellor of Germany in 1933. Admittedly, Hitler did not have majority support in Germany in 1933, but there can be little doubt that he would have had it later in the 1930s as the German economy recovered from the depths of the depression and his foreign policy appeared successful.

But democracies as well as dictatorships need leaders. For a leader makes things happen which hitherto appeared improbable — Britain’s survival in 1940, America’s recovery from the depression without sinking into dictatorship as so many European countries had done, the restoration of stable government in France after 1958, the founding of the state of Israel — all of these unexpected developments depended primarily if not wholly, on individual leaders — Winston Churchill, Franklin Roosevelt, Charles de Gaulle and Chaim Weizmann.

Of course leadership can be malign as well as benign — without Lenin, there would probably not have been a Soviet revolution, without Hitler there would probably not have been a Second World War. Had Stalin, Hitler and Mao never existed, many millions who perished would have survived. It is estimated that 60 million died in the Second World War, while at least 2 to 3 million, and probably more, died as a result of Stalinism, and at least 30 million died of famine between 1959 and 1961 as a result of Mao’s Great Leap Forward.

Leadership, therefore, is a neutral quality. It can be used either for good or for evil. Most leaders feel irritated by constraints. A main problem for democracies is to ensure that leaders are constrained from doing evil, while not being constrained from doing good. How can they ensure that their leader is a Franklin Roosevelt rather than a Donald Trump? For to give someone the power to do good may also give them the power to do evil. Still, many today bemoan the absence of courageous leaders in the democracies. Where are the modern equivalents of John F Kennedy, Margaret Thatcher, Charles de Gaulle and Konrad Adenauer?

But democracies do not need strong leaders at all times. In normal times, strong or charismatic leaders may be at a discount. A Harold Wilson, a Calvin Coolidge, a Francois Hollande or an Angela Merkel will suffice. Some countries — mainly those on the periphery of world affairs — Switzerland, New Zealand, Denmark, Norway, Sweden — hardly ever need strong leaders and are suspicious of charismatic leaders. I wonder how many can name the president of Switzerland. ‘Unhappy is the land that breeds no hero’, declares one of the characters in Bertolt Brecht’s play, Life of Galileo. But Galileo replies, ‘No — unhappy is the land that needs a hero’.

Democracies which play a much larger part on the world stage are also suspicious of strong leaders. Churchill, for example, would never have come to power had it not been for the desperate circumstances of 1940. In 1938, it seemed that after nearly forty years as an MP, he had hardly any political support. He saw himself as a failure, as, in the words of the influential newspaper proprietor, Lord Beaverbrook, a ‘busted flush’. In 1938, he told a friend, ‘My opportunity has passed. I am going to leave public life’. Oddly enough, one political leader with whom Churchill was later to have to deal, took a different view — the Soviet dictator, Joseph Stalin. In 1932, Lady Astor visiting the Soviet Union, told him that Churchill was finished. Stalin disagreed — when Britain was once again in trouble, he said, the old warhorse would be summoned back once again; and in 1931, the writer Harold Nicolson had said: ‘He is a man who leads forlorn hopes, and when the hopes of England become forlorn, he will once again be summoned to leadership’. And so it proved to be. But, even after the war, Churchill won just one of the three general elections he fought, the election of 1951. And that victory was won through the quirks of the election system. For Churchill’s Conservative party gained more seats than their Labour opponents despite winning fewer votes than Labour. Like Churchill, De Gaulle was also rejected after the war, retreating in exile to Colombey-les-Deux-Églises, to be recalled to power in 1958 only when he seemed the sole alternative to civil war or a coup by disaffected generals.

Some have suggested that the mature democracies have been becoming more presidential. The parliamentary system remains, it is suggested, but has come to be animated by a presidential impulse. In Britain at least, that view is open to very grave doubts. In 1970, Edward Heath’s Conservatives won the election though the ratings of Prime Minister, Harold Wilson’s were higher, while, in 1979, Margaret Thatcher’s Conservatives won though her ratings were lower than Labour’s James Callaghan. In more recent times, admittedly, both Margaret Thatcher and Tony Blair have been regarded as strong leaders. But both developed their leadership qualities after being selected as party leader, not before. The truth is that most advanced democracies are more frightened than inspired by strong leaders except during periods of upheaval.

Most journalists and commentators believe that the personality of the leader matters a great deal in election campaigns. But most political scientists do not share that view; and indeed psephological studies confirm that strong leadership is not a particular asset in general elections. The locus classicus on this subject is a comparative work edited by the late Anthony King, who was Professor of Government at Essex University, entitled, Leaders, Personalities and the Outcomes of Democratic Elections, published in 2002. King’s conclusion, based on a consensus of the views of political scientists, is that if there is a leadership effect at all, it is small and rarely decisive.

We may take as a case study the 1992 general election in Britain which the Conservatives unexpectedly won with a narrow majority of 21 seats. Labour’s Neil Kinnock believed that it was his personality that had prevented his party from winning the election. ‘One of the reasons voters put their crosses by the Conservative candidate was this innate feeling among a relatively small number of people that they couldn’t see me as prime minister. It’s just there in the biochemistry, as it were. It’s a pity but it’s a fact of life I recognise’.

Of course, in a close election, such as that of 1992, any of a host of factors can be responsible for success or failure. But Kinnock was almost certainly doing himself an injustice. Labour’s failure in 1992 can be perfectly well explained without bringing in his personality. Analysis of the election outcome showed that Labour had simply failed to convince aspirational voters in the south of England that the party was on their side. There is no need to introduce the personality factor at all. Indeed, and quite remarkably, Anthony King believes that ‘net leader effects actually helped the Labour Party marginally and harmed the Conservatives’, so that if Kinnock had been the Conservative leader and John Major, the Conservative Prime Minister, had instead been leader of the Labour Party, Labour would actually have done worse!

A similar general conclusion pouring doubt on the idea that leaders are fundamental to success in general elections can be found in a volume by Kees Aarts, Andre Blais and Hermann Schmitt, entitled Political Leaders and Democratic Elections.

Before Margaret Thatcher and Tony Blair, the greatest electoral landslides in the twentieth century in Britain were won by leaders who were far from striking personalities — Sir Henry Campbell-Bannerman for the Liberals in 1906, Stanley Baldwin for the Conservatives in 1924 and 1935 and Clement Attlee for Labour in 1945.

But, so it may be argued, since then the advent of television and then social media have altered things. We have moved from a primarily verbal culture to a primarily visual culture. This visual culture gives us the illusion that we know our political leaders even though we have never met them. Early evidence for the importance of charisma is supposedly to be found in the Kennedy/Nixon debates of 1960. As is well known, those who heard the debates on the radio believed that Nixon had got the better of the argument, while those who saw them on television believed that Kennedy had won. Yet it is often forgotten that the Democrats were at that time the majority party in America, and that Kennedy ran behind the congressional Democrats, not ahead of them, winning the presidential election by a whisker. What is surprising is not that Nixon lost, but that he so nearly won. There is little evidence that Kennedy’s charisma was a key factor in his success.

In the television debates before the 2010 election in Britain, many thought that Nick Clegg, the Liberal Democrat leader, was the most charismatic of the three party leaders, a view confirmed by survey evidence. Yet his party, although improving slightly on its performance in the previous election in 2005 in terms of votes, failed to achieve the electoral gains it had hoped for or to make a breakthrough. Indeed it actually won five seats fewer in 2010 than in 2005. It secured a smaller percentage of the vote than the predecessor party to the Liberal Democrats, the Liberal/SDP Alliance, had done in 1983, even though the Alliance had then been led by a distinctly un-televisual politician, Roy Jenkins. Survey evidence shows that the British public are wiser than the pundits give them credit for. The majority view was that Clegg was the nicest of the three party leaders, but not someone who voters could see as a credible or competent Prime Minister. He was likeable but might not prove effective in office. There is little evidence to suggest that his performance on television shifted many votes in the direction of his party, the Liberal Democrats.

But the conclusion that leadership may not affect election results should not be interpreted to mean that the leader is of no importance at all, only that the leadership effect is not to be explained in terms of personality or charisma. In 1997, for example, Tony Blair proved an effective leader of the Labour Party not because of who he was, but because of what he did — removing Clause 4 which had committed the party to the nationalisation of the means of production, distribution and exchange, distancing Labour from suspicions that it was a tax and spend party, that it was soft on crime, and that it was a largely owned subsidiary of the trade union movement. Margaret Thatcher proved an effective leader because she was able to convince voters that the Conservatives had a solution to the problem posed by predatory public sector unions, and, more broadly, to the problem of economic decline. Edward Heath had won the election in 1970 despite the fact that his ratings were lower than those of his opponent, Harold Wilson, because he had convinced voters that his Conservative party had the answer to rising prices.

Contrary to the view of most commentators, who are, by nature, political junkies, most voters spend little time thinking about politics, and want to be bothered by politicians as little as possible. They are interested less in whether a politician would be a personable dinner companion or has the qualities of a film star than whether she is able to meet their concrete hopes and aspirations, whether she charts a direction that they are prepared to follow.

In parliamentary systems, the outcome of general elections depends not so much on who the leaders are but on what they do. For, although talk of the presidentialisation of politics is exaggerated, the head of the executive is indeed in a quite different position from its other members, both in presidential and in parliamentary systems. The position of the person at the top is unique. The buck stops here was the sign that Harry Truman kept on his presidential desk. What he meant was that the decisions that came to him were all difficult ones. The easier decisions could be made at lower levels of government. So leadership in the sense of being a good decision maker is of fundamental importance in government., but it is often exercised successfully by those who are far from charismatic – Harry Truman himself, Konrad Adenauer in Germany, Tage Erlander in Sweden.

The founding fathers of the American constitution were concerned to impose checks and balances on the executive, to guard against the danger of an unconstitutional executive such as they believed George III to have been. The question then arose of whether such checks and balances were compatible with vigorous and effective government, an issue confronted by Alexander Hamilton in the 70^th of the Federalist papers. Hamilton was worried lest republican government produce a feeble executive, which ‘whatever it may be in theory, must be, in practice, a bad government’. Indeed, during the interwar years in Europe, dictatorships arose not because democratic governments were too strong, but because they were too weak. In France, for example, republican government in the 3^rd and 4^th republics was, in the words of the philosopher, Raymond Aron, ‘so afraid of great men that it was forced from time to time, to have recourse to saviours’ – Doumergue in 1934, Petain in 1940 and de Gaulle in 1958. The American law professor, Thurman Arnold, thought it an ‘absurd idea’ to imagine that ‘dictatorships are the result of a long series of small seizures of power on the part of a central government.’ Instead, ‘every dictatorship which we know flowed into power like air into a vacuum because the central government, in the face of a real difficulty, declined to exercise authority’.

In his message to Congress in July 1861, Abraham Lincoln asked, ‘Must a government of necessity be too strong for the liberties of its people or too weak to maintain its own existence?’

That is a question which the democracies have still to answer.