I got called up last Wednesday to ask if anyone at the Adam Smith Institute would go on the Daily Politics to explain why the Bank of England should raise its base rate (not exactly in those words). The producer was familiar with common free market ideas that argue that artificially low interest rates are blowing up a housing bubble which will later burst. I had to try to explain to the producer why I both agree and disagree with these sentiments: low interest rates do underlie economic limbo, but raising the base rate is not a solution and may produce yet lower real interest rates where it matters—throughout the economy.

The problem comes from the dual use, in the popular economic press, and even by top economists, of the term "interest rates" to mean both the stance of monetary policy and the cost of borrowing. This is understandable because during the Great Moderation of 1992-2008 all the world's most important macroeconomic authorities attempted to control the overall economy through adjusting one or a small number of key interest rates to achieve a consumer price inflation (CPI) target. At the same time, we are familiar with interest rates through our normal life: on loans, mortgages, savings, credit cards and so on. But acting as though the Bank of England directly controls these rates when it adjusts policy seriously obfuscates how the macroeconomy works and contributes to a lot of sloppy thinking.

Whereas the Federal Reserve has always used a form of quantitative easing (QE) to adjust a market interest rate—the Federal Funds Rate—the Bank of England has typically adjusted its base rate, which it calls Bank Rate, instead (updated). Bank Rate is the flat (nominal) interest rate it charges commercial banks for short term funding, and pays on their excess reserves. This sets a lower bound on overnight commercial lending, since it is always an option to lend or borrow money at Bank Rate, and therefore it is included in some market contracts, like tracker variable rate mortgages. The current UK base rate is 0.5%, a nominal number which translates to a negative real rate, but secured loans charge more like 3% in nominal terms, unsecured loans 8%, and credit cards 10%.

So we've established that the Bank of England sets a lower bound on interest rates with its Bank Rate. And we've also established that Bank Rate affects some other rates directly, principally tracker mortgages. We might also expect it to affect other rates in the economy—for example a cut will "ripple out" through the economy, because all other things being equal, it is now cheaper for banks to borrow from the BoE and they will thus be more willing to do so. Economists call this the liquidity effect. They will thus be more willing to lend cheaply and less willing to borrow from savers. So one effect of lowering the Bank Rate is to directly lower some rates, put a lower lower bound on others, and make others cheaper.

However there is an opposed reaction. Lowering Bank Rate doesn't just make loans cheaper, but it increases demand. It does so by injecting extra money into the economy (from the extra loans), but more importantly by signalling to markets that it intends demand to grow faster and that it is willing to take measures (such as further lowering Bank Rate or boosting the money supply through a QE programme) to make sure this happens. This is why stock markets react so strongly to a (policy) interest rate cut—all businesses are worth a bit more because they expect higher total revenues over their future.

But if firms expect higher demand in the future they will in turn demand more investment funds to put into projects to service that demand. This means that cutting the BoE's base rate puts pressure on effective market interest rates in both directions. It is an empirical question which direction the overall effect goes in—but this means that the simple coincidence of low real effective interest rates out in the economy and a low, by historical terms, Bank Rate, shows nothing. It could be that the best way to raise interest rates out there in the economy is to cut the Bank's base rate, or, since it can't go much further now, print money to raise inflation (which would ceteris paribus cut the rate in real terms). Look at the graph above for an illustration of how the Fed's changes in their QE programme (the red line) and their Federal Funds rate (the dark blue line) don't produce big shifts in (real) market interest rates like corporate bond returns and 30-year mortgages.

So my view on low interest rates is complicated. I think the Bank should get out of the business of setting rates altogether, and vary the size of the monetary base to control nominal income in the economy. But if the Bank is going to use rates as its key policy tool, it shouldn't raise them when a recovery hasn't quite taken hold—it's uncertain whether it'll raise market interest rates, but it will certainly choke off the demand we need for solid growth.