There is far more cash out there than can be explained by National Income Estimates of Expenditure and Output. This is as true of the US as it is of SA.
A modern economy laden with old fashioned cash
A peculiar feature of the modern economy is just how much cash lies around. The demand for cash appears largely unaffected by the growing use of highly convenient alternatives to cash to pay a bill, or check out of a hotel restaurant or shop. The value of transactions processed by the banks has grown very significantly with the use of credit and debit cards. A still more important development is the use of an online transaction that transfers ownership of a bank deposit from one party to another with a few clicks on a computer.
The extraordinary demand for dollars circulating outside the US banking system has attracted renewed interest following a revised estimate of the greenbacks held outside the US. (See James Surowieki, “The Underground Recovery” in the New Yorker, 25 April and referred to by John Mauldin in his free weekly investment and economic newsletter, “Thoughts from the Frontline”, 27 April, John Mauldin. The revised estimates of off shore holdings came from Edgar Feige of the University of Wisconsin, a pioneer in the analysis of the demand for cash.)
After allowing for the 27% held offshore, this leaves about US$750bn of cash in US wallets, purses, under mattresses and in safety vaults. This is equivalent to over $2000 cash stored by every person in the US. The average American family that is hard pressed for cash at the end of every month will be surprised to know how cash flush they are presumed to be. But as with many other metrics, such averages tell us very little about the financial condition of the average family. The distribution of these extraordinary cash holdings is no doubt highly skewed to the right, with relatively few holders holding the bulk of the cash for their own good reasons.
The demand for cash and other transactions in SA
In SA a similarly extraordinary growth in the demand for cash outside the banking system has been recorded. The rapid growth in demands for cash has taken place despite the impressive advances made in the availability and use of alternatives to cash in SA, as in the US and elsewhere, with the adoption by banks of new technologies. The notes in circulation outside the banking system grew from nearly R23bn at the end of 1999 to over R81b by the end of 2012. That is at an average compound growth rate of 11.3% p.a. Adjusted for inflation the average compound rate of growth was 4.4% p.a. (See below)
According to the 2011 census there are some 15 milion households in SA. Dividing R8.1bn of cash in circulation by these 15 million households would mean that the average household in SA held on average as much as R5 400 in cash at the end of 2012. A small, fairly constant proportion of this cash will be held in neighbouring countries, especially Zimbabwe, but these demands are unlikely to account for more than 4% of the rand notes and coin currently in circulation.
In the figures below we show the strong growth in the value of electronic transactions effected by the SA Banks. Not surprisingly, given their convenience, the use of electronic fund transfers (EFTs) has grown significantly while the use of cheques has fallen away. The use of credit cards, while still relatively small, has more or less kept pace with the other forms of exchange since 2002, as we show below.
The value of purely electronic transactions facilitated by the banks grew from R155.2bn in January 2003 to R603.5bn in December 2012, that is at an average compound growth rate of approximately 13.6% p.a. This growth, while impressive, is only about 2% p.a faster than the growth in cash, despite the switch from cheques to EFTs. The average EFT processed by the banks is now about R8 583 per transaction and the average credit card transaction is of the order of R545. These average sized transactions are lower than they were in 2003 when adjusted for inflation.
The average low income SA household is too poor to hold this much cash at the expense of the inadequate food, clothes, energy or educational services they consume. And so we should conclude that the heavy lifting of cash in SA is probably not being done by the road side hawker or marginal retailer, but by more significant business enterprises and their owners.
How can we explain the demand for cash? Using cash escapes surveillance.
The question then is what is all this cash on hand being used for? Cash very obviously serves the interests of those who wish to hide their income from the tax authorities or the officials responsible for means testing welfare benefits. Using cash helps escape surveillance by government and the financial system generally. Honestly declaring extra income earned that takes welfare recipients beyond the income thresholds or tax payers into higher tax brackets can make for what are effectively very high rates of taxation of income at the margin. An extra $100 or rands of extra income earned and declared may well mean more than a hundred of sacrificed benefits or as much in taxes levied.
The currency approach to measuring informal activity
Clearly there is a lot more cash out there in use in SA and the US and in many other economies than can be explained by officially estimated incomes or expenditure, especially given the growth in the use of the alternatives to cash. The question then is just how much income and economic activity goes unrecorded when cash is exchanged for labour and goods and services? Just how much income or expenditure is going unrecorded because the transactions and the value they add are made in cash and are not reported in any reliable consistent way?
An estimate of how much activity is not recorded can be found by observation of the demand for cash itself. The essence of the approach is to attempt to explain and predict the demand for cash using incomes, prices and improvements in payments technology, measured as the value of electronic transactions processed by the banks, as explanations of the demand for cash. The observed higher demands for cash, the demands for cash that cannot be explained in this economically sensible way, then becomes a proxy for estimating the unrecorded levels of economic activity.
How much economic activity is not recorded?
The New Yorker article suggested that as much as $2 trillion worth of economic activity in the US may be going unrecorded. Given that the US GDP is officially estimated at just over $16 trillion, this would make unrecorded activity or the informal economy in the US equivalent to about 12% of the official US economy. Such an estimate is on the low side of estimates made for a number of developed economies, using a variety of methods to supplement national income estimates, including the currency demand models.
Many years ago I attempted to replicate the studies of Feige and others using SA data. In those days the SA economy and its labour market was severely infected by apartheid. In particular, the pass laws and other racially inspired laws and regulations prevented employers providing work and employees from offering their labour. These controls would have encouraged “illegal” activity and employment and the use of cash to avoid detection. My ball park estimates of unrecorded activity were equivalent to over 15% of the official economy. Or, in other words, the SA economy was then perhaps 15% larger than indicated by estimated GDP.
This approach did not find favour. The official estimates of unrecorded activity in SA to this day, officially assumed to be very largely informal retail activity, are estimated to be only about five per cent of recorded activity. This estimate is extraordinarily, perhaps unbelievably low, by international comparisons.
The incentives to use cash in the US and South Africa
For the US the major incentive to use cash may well be, as the New Yorker suggests, to avoid losing welfare benefits. In South Africa a further more important reason for using cash may be to escape not only tax or avoid the loss of welfare benefits but also and more importantly to escape the burdens of a highly regulated labour market. The incentive to use cash, rather than banks, to side step the regulations of the labour market and also by so doing to escape the supervision of the Receiver of Revenue, is surely a powerful motive for using cash.
More unrecorded activity means more unrecorded employment. If so this does not weaken the argument for a less regulated economy
If we are underestimating income and expenditure we are also underestimating actual employment. The only employment numbers we can be reasonably be sure of are the jobs offered by the formal sector. Unrecorded economic activity and unrecorded employment are therefore also matters of conjecture.
If SA has less of an employment problem than the official estimates indicate, given the unrecorded economic activity and associated employment, it still has as much of a poverty problem. The solution to SA poverty is faster growth, especially in faster growth of formal employment. But such growth in employment will not be realized or recorded unless the incentives for all businesses, especially small businesses, to operate formally are much improved.
Conclusion: Greater economic freedom for South Africa will add to incomes and employment and reduce the demands for cash and increase rather than reduce tax collected.
Encouraging formal employment and less evasion of taxation requires freer labour markets, less complicated income taxation, lower business income tax rates and much more sympathetic regulation of small businesses and their owners. Such steps might well raise more rather than reduce tax revenues as small businesses elect to operate above rather than below the radar screen. If they did so one of their attendant benefits would be access to a much more convenient payments system.
Progress in this regard may well be recognised by slower growth in the demand for and supply of cash. Unfortunately SA, with new licensing demands on all businesses that would seem to be in the interest only of the officials attempting to enforce new licensing laws, seems to be moving in the other direction. Brian Kantor