Implications for the Standard of Living
In a controversial study, David Mitch (1992) has argued that, during the nineteenth century, England was grossly overeducated. The stock of human capital was well in excess of what was called for by the jobs performed, and its contribution to the epoch's rapid economic growth was therefore small.
This enormous literacy surplus was due to two circumstances: a steady expansion of educational facilities and a rising demand for reading and writing skills as a form of consumption by an increasing well-off society. Put in other words, the average product of human capital investment was positive and quite significant but at the margin, it may have been much smaller as a result of the complementary, non-investment role it had in society.The present study detects a similar though not identical pattern in Europe during the latter half of the early modern period. At this time too, there was a surplus of human capital in the sense that certain segments of the population were educated without correspondence to their professional needs. In contrast, as we saw above, the total stock of human capital varied to quite a large extent in consonance with the
evolving requirements of the economy. This suggests two inferences. One is that the contribution of human capital to any growth there may have been will, in all likelihood, have been significant. The other is that, in this case, the literacy surplus was probably still quite small although it was on a long-term upward trend. Our interest lies in the second of these issues, that is, the extent of the acquisition of human capital as a form of immaterial wealth that individuals procured in order to enhance their spiritual well-being in particular and their welfare in general. The next question is whether this can be quantified and, if so, with what result.
To achieve this, we begin by treating the capacity to read and write (proxied by the ability to sign) as a costly asset in the portfolio of consumption durables that individuals accumulated and enjoyed during their lifetimes—furniture, clothing, ornaments, etc.
The assessment of this portfolio and its components corresponds, somewhat arbitrarily, to the end of their lives because that is when they were usually inventoried. As is common in the literature, the total value thus obtained is employed, after suitable deflation, as a yardstick for gauging the standard of living and the relative weight of the items contained in this basket of more or less durable goods (Schuurman and Walsh 1994). In the present case, it serves for appraising the relative importance that human capital had in the overall picture. The first problem then is to put a monetary figure on human capital and this can be done in one of three ways.Unlike most consumer durables, which would have had some sort of second-hand use and therefore a potential market price, human capital disappears with its possessor. In this perspective, the correct approach would be to attribute a value of zero to it. Since, however, what we seek to capture with this exercise is the value of the asset's fruition during its owner's lifetime, this solution appears unhelpful. A second way is to value the stream of utility over a given period by a conventional measure such as the asset's rental value, using normal depreciation and discount rates. Given that human capital can be assumed to suffer no depreciation—it might indeed appreciate with use—this should constitute a relatively simple operation once we have its capital cost. The difficulty this time lies in estimating the equivalent stream of satisfaction generated by all the other assets in the inventory, an elusive target given that in many cases we would not know, either their current age or their depreciation schedules. The third option is a stock, rather than a flow solution, and it is this we follow here. It consists in taking the historic cost of human capital and confronting it with the current market value of the material assets in the portfolio. The disadvantage of doing this is that we shall be comparing articles valued by different methods.
This is mitigated, however, by the fact that human capital, in principle, does not depreciate and so it matters little when it is appraised. Furthermore, once it is acquired, it cannot be transacted, which means in effect that there is no other practical way of pricing it.The present estimate refers to early eighteenth-century England and France, two cases for which reasonable data exist with respect to all the required parameters and where a consumption revolution is supposed to have been in full swing. The historic cost in question aggregates two items: the cost of education and the opportunity cost
incurred by withdrawing the learner from the labour market. As regards the former, contemporary observers noted that normally it took three years to learn to read and, following that, another three years to learn to write (van Deursen 1991), though there are numerous instances on record of gifted individuals who required only a few months for either (Spufford 1979). Since we are considering only those who could sign their names but could not necessarily write any more than that, we assume that the second part of their education was limited to an additional year. This gives us a lower bound estimate. At mid-century, in Birmingham, a low cost elementary school charged pupils 3—5 shillings each per quarter (Money 1993), that is, taking the lower value, a total of £2.4 for the four year education of our premise and very close to Mitch's (1999) figure of £2.0 for the early nineteenth century.36 In France, monthly fees varied between 3 and 4 sols for tuition in reading and between 4 and 6 sols for learning to write (Houdaille 1977; Grevet 1987). On an assumption of ten months a year of schooling, this would entail a global cost of elementary literacy of an order of 6.5 to 9 livres tournois. Since Chartier et al. (1976) put a figure of 13 livres on this, we adopt for the present exercise an intermediate estimate of 9 livres.37
Putting a value on the labour time lost in the process is a good deal more hazardous and depends on several not easily specifiable factors. To begin with, there was a wide range of possible ages over which education could be acquired.
We assume here, again for the sake of a lower bound estimate, an early, low opportunity cost range, from 6 or 7 to 10 or 11 years. In the second place, there is the vexed question of the role of child labour in the pre-industrial European economy, on which views are divergent while little hard evidence is available to help sort them out. In proto-industrial areas, children were far more sought after for paid employment but in the majority of situations, which were agrarian based, occupational opportunities for them were scarce. Boys and girls became farm servants usually at thirteen or fourteen years of age but smaller children helped in the family and were unlikely to be hired for any significant task or length of time. Around 1800 in England, Horrell and Humphries (1995) have established a participation rate in agriculture for the under tens of only 15%, a state of affairs which is confirmed by Cunningham (1990).The problem is compounded by a dearth of information on wage rates, which are available only for adults. As an upper bound limit, we take a child's wage to have been one-fifth of that of an adult (Spufford 1985; van Deursen 1991) and assume an average annual occupation rate for them of 15%, in other words, 45 days of employment. The lower bound limit is given by a situation of full unemployment and hence zero income. Between 1700 and 1750, an unskilled male rural worker in early eighteenth-century England earned about £20 a year (Mitchell 1988), and therefore the opportunity cost of achieving proficiency in reading and writing would have been something like £2.4 at the most and, at the least, nothing. The full historic cost of being literate in this context thus amounted altogether to between £2.4 and £4.8. A similar exercise for France adopts the same assumptions regarding time of work and child/adult wage ratios. In the early eighteenth century, the remuneration
for unskilled labour in the countryside varied between 10 and 15 sols per day (Morineau 1972; Chartier et al.1976; Baehrel 1988) and if we take 12.5 sols as representative, the opportunity cost of a child acquiring literacy would come to 22.5 livres and the full cost, including schooling, to between 9 and 31.5 livres.
Having quantified the value of human capital, the next step is to place this in context, the most appropriate one being naturally the stock of wealth owned by the people under consideration. In early eighteenth-century England, rural labourers left estates worth on average £16, while slightly up the social ladder, for husbandmen and small farmers, the figure was double this amount (Weatherill 1988). This included all goods and chattels but the average value of household goods alone, the reality on which the standard of living discussion focuses, was far lower—respectively £5 and £8.38 The implication is that even at the lowest end of the estimated range the cost of acquiring the most basic sort of literacy was substantial in terms of the domestic economy of the social strata in question—somewhere in the region of 50% of the accumulated material wealth of individuals at the end of their lives. For France, the contrast between household wealth and the capital cost of literacy is less sharp but still implies a recasting of the standard of living evaluation based exclusively on material goods. The inventories of rural labourers that have been examined for this period were characteristically below 500 livres, and their household possessions did not exceed 100 livres in total so that here too it can be said that human capital represented a valuable asset in the lives of a great many of the humblest people who acquired it (Baulant 1975; Waro-Desjardins 1993; Boehler 1995).
The results of our estimations are of the roughest kind and the only claim that can be safely made on their behalf is that they provide us with an order of magnitude for the importance of a critical and increasingly significant immaterial good in people's lives. One finding this points to is that if human capital is treated on par with material consumer durables then the level of welfare in the eighteenth century for the mass of the population has to be revised upward relative to earlier centuries.
A second is that the spread of literacy during the 1700s, in particular for those on the lowest rungs of the social ladder, clearly indicates an improvement in terms of the broadly defined standard of living that has been proposed above. Not only is the notion of a consumer revolution hereby reinforced but also it becomes clear that, in fact, this process reached much further down, into the realm of the poor and ζun-inventoried,, whose welfare status on the whole has escaped the scrutiny of the historian. Finally, it should be noted that human capital also enriched men and women in other reaches of society, and, among the better-off, it seems likely that this immaterial good would have had a higher value for the inventory since their schooling was better. It seems likely, however, that in the case of the humble, the weight this represented in total wealth would have been greater, a fact that suggests a further possible revision—the gap between the rich and the poor may have been less than has been thought until now.What is more remarkable, though, is that this numerous group of poor literates spent their resources to obtain this capacity, knowing full well that they were not
investing in a producer good. For them, this human capital could serve practically only as an end in itself to generate a stream of welfare of a wholly immaterial nature. Although our assumptions may be faulted to some extent and our quantification, albeit a lower bound result, stands to be corrected, arguably the substance of our inference is not affected. The fact remains that taking human capital into consideration in this way is advisable and liable to affect the traditional calculation of standard of living levels to a substantial degree.
6.