There is something of a brouhaha rising up in the United Kingdom over Prime Minister Theresa May’s executive pay reforms. Labour leader Ed Miliband is claiming they were his ideas; the Institute of Directors thinks they are too harsh (so do a lot of other industry types); and the press and the public are baying for fat cat blood.

In the United States, we have the prospect of a Trump administration rolling back much less draconian regulations surrounding pay; in England, totally the opposite.

But what happens when government gets involved in executive pay? The answer is: always the opposite of what it wanted.

So, what’s going on?

May has a whole set of proposals on introducing more binding pay votes, publishing a CEO/worker pay ratio—a likely target for a Trump “wrong” before it even gets going in the United States—more disclosures on annual bonus targets, simpler pay structures, committees of shareholders and/or stakeholders to advise on pay, more engagement, and more experienced remuneration committee chairs. Like other groups, May doesn’t really have anything to suggest in the way of simpler pay structures, but rather asks the question—does anyone else?

But it’s not just the government; there has been an endless string of advice from all sorts of groups, including the Executive Remuneration Working Group, the Investment Association, shareholders, and industry groups. And, most recently, The Purposeful Company Interim Executive Remuneration Report from think tank the Big Innovation Centre. This says that shareholder guidelines and the U.K. Corporate Governance code should enable companies to adopt simpler pay structures that de-emphasise short-term cash incentives and spread the money to longer-term share schemes. One is tempted to ask: What’s stopping them now? It also says companies should publish a fair pay charter and consult with employees on changes in the CEO’s pay and those of the workforce generally—but not to publish a single year pay ratio. Reporting regulations should be updated so people can actually understand remuneration reports. And a binding vote regime should only be introduced after two significant protest votes.

Andy Haldane, the Bank of England’s chief economist, was among many who rejected May’s proposals for binding shareholder votes on pay and the publication of salary ratios. On the other hand, the Labour party is wondering why she is stealing its own ideas. A Labour party manifesto from 2015 called for employee representation on boards; and as far back as 2011, leader Ed Miliband was calling for the publication of the CEO/worker pay ratio.

But what happens when government gets involved in executive pay? The answer is: always the opposite of what it wanted.

At the same time as all of this, a survey from the Pensions and Lifetime Savings Association (PLSA) found that 87 percent of pension fund managers say pay is too high. Even worse, an equally significant 60 percent say high levels of pay at the asset managers they hire, the people actually voting on pay, is “what is preventing them from properly holding companies to account on the issue.”

Then IoD Head Simon Walker was asked to speak by the High Pay Centre—a group against high pay, not for it—and he responded with a speech that warned that after “Brexit and Trump, business should expect a new level of scrutiny and questioning of their role in society.” On high CEO pay, he said: “Unless boards show that they are listening, and responding, to the mood of the times, the Government’s trigger finger will just get itchier and itchier.” But he seems to agree with me that regulation is not the way, since he sings the praise of Labour MP Ed Balls: “He was against regulating executive pay and he believed that was true of most MPs on both sides of the house. But public and media pressure can make most politicians intervene on topics they knew were no business of government.”

As anyone who has read anything I have written over the last 28 years knows, I am not a fan of oversized pay packages, but I am also not a fan of government intervention in pay packages. As I said earlier, regulation on pay structures tends to have not just unintended consequences but absolutely the opposite effect of what it wanted.

Section 162(M) of the U.S. tax code that sought to limit fixed pay to $1 million ended up raising all base salaries to $1 million. By shifting everything into bonuses and equity incentives, Section 162 (m) is also widely seen as being the direct inspiration for the huge inflation in U.S. CEO pay. Limits to golden parachutes ended up making three years’ salary and bonus a floor for severance rather than a ceiling. Rules meant to limit Wall St. pay at the time of the financial crisis led to massive increases in base salary which—as the building block of pay—then led to massive increases in total pay once the limits were dropped. And in the United Kingdom, the rules on bankers’ pay—the ‘allowances’ that banks were allowed to pay in order for annual bonuses to be reduced—just led to everyone reducing bankers’ bonuses and then replacing that amount with the maximum allowances accepted; so now that pay is fixed, rather than based on performance. Massive increases in disclosure regulations are accused of leading many CEOs to ask “how come he gets that much; why can’t I?” thus more pay inflation. Publishing pay ratios might even do the same.

Not good results, are they? Regulations: 0, CEO pay: 7. So what is the solution?

Any solution is going to have to come from outside corporate boards, because I think we can safely say that neither boards nor CEOs will voluntarily simplify and reduce pay levels.

But who’s left? Investors?

They own the companies and should have more say in how pay is set. After all, it’s their capital that is being used to pay CEOs and they should have an interest in what kind of return they are getting from that capital.

But are mandatory say-on-pay votes likely to have a beneficial effect? Figures in the government’s own green paper show that most shareholders don’t vote against fat pay packages. In fact, a sizeable minority don’t vote at all. The green paper has some suggestions for solving this, but since there have been less than a handful of mandatory votes against pay, changing that rule is hardly likely to have a sweeping effect. There are some responsible shareholders out there who regularly take a stand against excessive pay, but they are a small minority. And as we have already seen from the PLSA, high pay among asset managers might make them reluctant to vote against high pay.

So, not investors, then.

What’s really needed, I believe, is the introduction of some perspective into the boardroom. Someone who can ask the question: “You’re making how much a day?” without someone being able to turn around and say: “Isn’t that about what you’re making?” Whether that means employee representatives on boards, customer representatives on boards, or both, or May’s own stakeholder committees advising and consulting on pay—instead of executive pay consultants—is not important. Whatever form it takes, perspective is what is needed.