Jonathan Portes offers an insider's perspective on the debates that preceded Black Wednesday, 1992, and the economic lessons that can be learned.

The economic and political history of the UK's ill-fated experiment with the exchange rate mechanism (ERM) has been picked over many times, and the broad outlines of the story are well known (here's a range of well-informed views). So this is more of a personal recollection of how it felt at the time.

I spent the night and early morning of 9 April 1992 - the general election - in front of the TV, with my then boss, Jeremy Heywood, now the cabinet secretary, and our friend Suma Chakrabarti, now running the European Bank for Reconstruction and Development. As the result became clearer, Suma and I got drunker, and more voluble; Jeremy chain-smoked and remained as impassive as ever.

So when I arrived at the chancellor's office at the Treasury the next morning, I was less than usually tactful (yes, this is in fact possible). The first thing I said to Norman Lamont (who was, I suspect, equally hungover) was, 'I wasn't really expecting you to be here today.' To which he replied, 'Me neither.'

And so we returned to work (I was Norman's speechwriter). While the Conservative party as a whole was understandably euphoric at their surprise victory, the elation didn't last long for those responsible for economic policy. The economy was also suffering from a severe and much longer-lasting hangover. The hubristic tax cutting of Nigel Lawson and the financial sector excess of the late 1980s had led to a surge in inflation and a rise in interest rates; in 1992, we were still bouncing along the bottom of the ensuing recession. If all this sounds vaguely familiar, it should.

In May, the immediate problem was obvious. From a domestic point of view, the appropriate level of interest rates, given weak demand, was much lower than that necessary to maintain sterling's position in the ERM. Moreover, it was becoming increasingly clear that sterling was overvalued; even in the depths of a recession, we still had a large current account deficit. However, the government line - and Treasury's line - was that there was no alternative; things would eventually get better.

I didn't believe it. My willingness to argue a different line to the entire senior Treasury management was probably a mixture of sound instinct and youthful arrogance, and I was sufficiently unimportant that I could just about get away with it. Jeremy and I wrote a short, private paper for the chancellor. We argued that the fundamental problem was that we'd joined the ERM at the wrong rate; sterling was overvalued, meaning that we were stuck with a structural current account deficit. The only way to maintain the peg would be through what is now, in the eurozone context, referred to as 'internal devaluation' - that is, setting real interest rates at a higher rate than dictated by internal conditions - and a long and grinding squeeze on wages and prices.

Our solution? We didn't dare suggest complete abandonment of the ERM. One possibility was for sterling to 'realign' - that is, to devalue - to a considerably lower rate, thereby boosting exports and allowing interest rates to fall. Even better, politically and perhaps economically, would have been if the Germans could have been persuaded to realign upwards, so avoiding the perception that sterling was being singled out; but the French were resolutely opposed to any devaluation of the franc.

Could either route have worked and so kept the UK in the ERM? In restrospect, probably not - it is very difficult to 'manage' a devaluation, because no one knows the 'right' rate. An attempt at managed devaluation would probably have led to unmanaged exit. So a final, more radical but perhaps more plausible option was to leave temporarily, allow sterling to find its own level, and then rejoin when both the currency and interest rates had stabilised.

In any case, none of this was going to happen. Norman called us in. He said (I paraphrase): 'I've read this. Broadly, I agree. But forget it. What I want to do instead is to make a speech setting out the best possible case for the UK's membership of the ERM.'

I started writing, and together we wrote what Norman and I both agreed was easily the best speech we produced during his time as chancellor. It set out, carefully and logically, the reasons why ERM exit would be a disaster for Britain. The UK had long suffered from periodic cycles of boom and bust, most recently exemplified by the deep recession of the early 1980s and the unsustainable boom of the late 1980s. Monetarism - targeting various measures of the money supply - had failed miserably. The alternative was to 'import' credibility from a country with a demonstrated record of maintaining low inflation while avoiding boom and bust - Germany - and we could do exactly that by tying our exchange rate and monetary policy to theirs.

And the early experience of the ERM was relatively positive; our interest rates did converge somewhat with Germany's at the same time as inflation fell. Leaving now would be a disaster: without the discipline of the ERM as a monetary anchor, wage pressures would rise and so, in turn, would inflation. This would result in higher interest rates, both shorter and longer term, choking off recovery. The consequent fall in sterling would further exacerbate inflationary pressures. Nor would there be a rapid return to growth; you couldn't devalue your way out of a recession. In the longer term, we'd be back to boom and bust.

Almost all of this turned out to be wrong, of course, but it was a clear, coherent and - we thought - convincing statement of the government's case. In domestic political terms, it had the desired effect. In the international financial markets, however, things were gradually slipping out of control.

So what happened next? Well, the high-pressure atmosphere of private office was beginning to burn me out. I didn't think I'd ever write a better speech. And somehow I'd managed to blag the Treasury into sending me to Princeton to belatedly learn some economics. So by the time of 'Black Wednesday', I'd moved to the US.

I did see Norman shortly thereafter. We were both in Washington DC, and we drank champagne on the embassy terrace and discussed that day's referendum on the Maastricht treaty in France (51-49 for 'yes', if you've forgotten). He was cheerful, although well aware of the impact of ERM withdrawal on both his credibility and that of the government as a whole; this was the night before his famous 'singing in the bath' remark to the British press.

And indeed, liberated from the constraints of a policy that he did not believe in, Norman's macroeconomic strategy was exemplary. He moved quickly to introduce an inflation targeting regime, pioneered in New Zealand and Canada just a couple of years previously. It was extremely successful, establishing credibility remarkably quickly, and prepared the ground for full Bank of England independence.

The success of fiscal policy in the 1990s, under first Norman and then Ken Clarke, is less well known. But, as I have argued elsewhere, he made the right decision for the right reasons, in sharp contrast to George Osborne, 20 years on. Policy was tightened sharply, but only after recovery had been firmly established, rather than, as now, before that gain was in place. While there are of course a number of reasons why the post-2009 experience has been so different from the post-1992 one, despite a similar devaluation, the failure of the current chancellor to learn from the success of his Conservative predecessor is clearly one.

What did I learn? Well, there are numerous economists who, if you asked them what Black Wednesday implies for the conduct of monetary and exchange rate policy, could happily start a fight in an empty room. But for me, the most important lesson was a more general one about 'credibility'- a concept often used and abused by both politicians and economists. As with the ERM, the argument made by the current government and its supporters for sticking to its fiscal consolidation plan, despite its evident failure, is that the strategy has established 'credibility', especially with financial markets, which can only be preserved by sticking with it.

But of course this is not a justification, economic or otherwise, for the policy itself. Instead, it is an argument for never changing policy at all. It relies on an odd view of market psychology, one that says markets have more confidence in governments that never adjust policy, even when it is sensible, from an economic perspective, to change course. But, as Black Wednesday (and numerous other examples from economic history) show, this is not plausible. Markets can, of course, be irrational. But there is neither a theoretical reason nor any empirical evidence to suggest that they are irrational in this particular way. The real hit to credibility comes from sticking to unsustainable policies; and economic success comes from abandoning them and doing something sensible instead. That is one lesson from Black Wednesday we could usefully remember.