1. Inequality and Economic Growth
Amidst Piketty fever it is useful to consider not just the issue of whether wealth become more concentrated over time but also the issue of whether this is economically harmful.
For the sake of the model below we will assume that wealth does become more concentrated over time. Actually I think this is more complicated and whether R>G depends on the state of the asset price bubble cycle – but the trend over the long run is for R>G.
With a steady rise in rentier income there may be a gap between the budgets of consumers and the price of fixed consumption assets (e.g. land on which homes are built) as those assets become ever more subject to oligopoly pricing and economic rent. This may require consumer to take out loans to up their budget constraint. But this by itself does not explain why this must harm growth. The loanable funds argument – debts = credits, advanced by Bernanke, Krugman etc. to state why this does not matter may have been crushed by the endogenous money crowd but the fact that there can be a net change in spending power from debt dynamics does not in itself explain why this must be economically harmful.
If there is a net increase in debt then by definition R will be increasing faster than G as those benefiting from capital accumulation don’t need to be so heavily geared in their investments. But in terms of economic growth this is begging the question – is concentration of wealth harmful or not.? Yes it is and I think that there is a simple reason why.
2. Rethinking the Budget Constraint
I stopped blogging on deep economic theory issues here for a while partly because of work pressures in a new job in a new country but partially because the issue of the relationship between liquidity, prices and money velocity seemed undefined at the most basic level in economic theory, though Keynes had made a good start in shattering classical assumptions. Working this through seemed like a job of a couple of years in hermit like retreat in double entry tables and continuous time equations. But there is a simpler way.
Think of the way neo-classical economic defines the ‘budget constraint’ – it is atemporal and deeply wrong.
A budget constraint represents all the combinations of goods and services that a consumer may purchase given current prices within his or her given income. (wikipedia)
Then the choice between two goods is defined by an indifference curve. Bread, milk what ever.
However when it comes to large consumption assets we cant simply define a linear bread, house purchase whatever curve.
The underlying assumption seems to be that the budget constraint is a flow, at the beginning of this month this becomes a stock and we spend up to that stock constraint, live off what we have bought and what we don’t spend is our liquidity till the end of the next month, a buffer for unforseen expenses. Indeed many formal DGSE type models explicit model this as a ‘cash in advance’ or similar constraint. People don’t spend unless they have ‘cash in advance’
But spending on big ticket items is not like that – we have to save (spend less than income) for them. Lets say you are renting a flat not even buying in a new country. You will need both the month, or several months in advance plus a deposit plus the cost of temporary accommodation whilst you look for the flat. Imagine this ‘indifference curve’ between staying in a hotel and renting the flat. You will need to be able to accumulate liquidity sufficient to move (and reduce the rate of spending on housing services) but you cannot until you have sufficient liquidity to move – which requires a conscious decision to not spend to save for future purposes. Nor is it an an indifference curve between spending now and spending in the future, as there is no option to spend no until you have the liquidity to do so. Liquidity is always prior in time to indifference – and to not appreciate this is to commit the ergodic fallacy. This is at the core of Keynes fundamental point of demand being logically prior in time to supply.
We need then to redefine both saving for future consumption and the budget constraint in rigorously stock flow consistent terms. We also need to be very careful when modelling in continuous time, as payments to income and spending from income for large consumption goods are discontinuous jumps not steady flows.
3. The Macroeconomic Impact of Reducing Spending to Pay for Goods with High Economic Rent
When we are accumulating funds to pay for higher olipopoiostic prices including an element of economic rent we reduce the velocity of money. If R>G then increased concentration of assets will increase economic rents. Consumers will have to save longer to buy the same large ticket goods like houses, cars etc, go to college etc. As they will have to defer spending for longer monetary velocity slows. It is really is as simple as that.
4. Potential objection – Money in Idle Balances is to Put to use by Banks
The first potential objection, one that Bohm Bawerck might have made, is that money in idle balances is not economically idle it is made use of by banks. Regular readers will know I agree with this. Money is not a simple + or -value – it is an asset liability pair, it must mathematically be treated as a tuple, and a fungable one which can be treated as an asset, even though it is accounted for as a liability, as long as future claims on that liability are honored. Lets say you were out shopping with a friend. They go into the hairdressers and say ‘look after my bags’ you then rush next door to the book makers and bet their money on the horse, knowing that if you lose you have money in your own pocket to meet the future liability. The horse wins, you honor the liability and keep the profit. Those who claim that liabilities can never be invested need to explain where this profit came from. Banks are like this – they gamble with your money and know that it is unlikely in their fractional reserve model that everyone will seek to reclaim their assets at once and all bets will fail at once. This is precisely the model of ‘liquidity transformation’ that Warren Buffet has used to accelerate his accumulation of capital by using insurance deposits as investments. Such an approach is entirely consistent with a endogenous double entry view of money creation – not only a crude loadable funds view – as many theorists going back 150 years have recognised (see my article here).
The objection might be that this ‘idle’ money is really circulating through this channel (which as I have argued increases the charter value of banks and their ability to lend – their lending power) and so does not reduce money velocity. Not so. Again accumulation and lending are not simultaneous they occur in time – there will be a delay between an increase in the power of a bank to create money ex nihilo and its creation. You could use the law of large numbers to argue in aggregate this would be very small as the acts of depositing and lending are occurring all of the time and the scale of any individual loan will be small to the capital base of the bank. True but only fully at the limit if there were one and only bank and an infinite number of savers, depositors and borrowers.
The second response to this is who is actually benefitting from the profit stream of this bank lending. The owners of bank equity of course, so if R>G the effect is simply reinforced.
The third response is that the higher proportion of wages spend on economic rent the greater the ability of banks to lend through liquidity transformation. Hence they must expand lending as if they don’t lend it out they will lose out to competitor banks who do. Hence they will push interest rates down. So interest rates will be lowered. The primary driver of interest rates is not time preference but the distribution of wealth. The high level of credit creation of debt in highly unequal societioes leads to a boom and bust culture and massive unnecessary capital destruction during the bust phases.