Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
Coin Evidence as a Proxy for Wage Labour in Medieval England and Wales
Spatial and Temporal Changes, 1066-1660
Preliminary short paper – ESSHC 2018
Rombert Stapel, International Institute of Social History
– Please do not cite –
Abstract
The rise of wage labour is considered one of the indicators for the transitional nature of the
late medieval economy in Northwest Europe and the emergence of capitalism. It is not
surprising therefore, that wage labour has attracted the interest of many economic historians.
There has been particular interest in wage levels, creating long time series of wage levels in
various parts of medieval Europe. Regional variations of these wage levels are often studied
to research (labour) market integration.
While such time series are incredibly informative, they are not an indicator of the
number of people that worked for wages in a particular area, and to what extent wage
labourers were fully or partially dependent on the income from wages. Recently, an
alternative indicator was proposed for how many people worked for wages in a certain
country. This indicator is the level of small denomination coins per capita, worth a daily wage
or less.
In this study, I will examine to what extent small coin evidence, both archaeological
coin finds as well as coin production figures, can be used to study both spatial and temporal
variation in wage labour in late medieval England and Wales; and which intermediary steps
R.J. Stapel – Coin Evidence as a Proxy for Spatial and Temporal Changes in Wage Labour
ESSHC 2018
2
are necessary to come to such an evaluation. The spatial component is added at the end of
the paper and will be addressed in the presentation in more detail. The paper below is meant
as an accompanying piece that contains the steps taken to come to detailed monetisation
levels in medieval England and Wales over the course of several centuries.
Introduction
The idea behind using the availability of coins as a proxy for the presence of wage labour is
quite simple and elegant, and was further developed in recent years by Jan Lucassen.1 It builds
on the premise that the availability of ‘a substantial stock of currencies per capita in
circulation, consisting of denominations equalling the value of one hour or less of waged work’
in particular areas of the world can be used as an indicator for the presence of wage labour in
those areas. In areas where the per capita stock of these coins of small denominations is five
to ten times the prevailing wage level of a skilled worker, Lucassen argues, a level of ‘deep
monetisation’ is reached.2 The advantages of this approach are twofold. First, it offers a new,
consistent method to compare different regions of the world over time and space. Second, it
can be used to study areas of the world for which little or no data is available on the size of
the group of wage earners in society.
While Jan Lucassen has applied the method to various countries or larger regions in
Europe and Asia, there is no restriction on the geographical level of analysis, as long as a
meaningful number of coins in circulation per capita can be calculated. In this paper, I will use
1 J.M.W.G. Lucassen, ‘Deep Monetisation: The Case of the Netherlands 1200-1940’, Tijdschrift voor Sociale en Economische Geschiedenis 11 (2014) 73–122; J.M.W.G. Lucassen, ‘Deep monetization, commercialization, and proletarization. Possible links, India 1200-1900’ (Noida: V.V. Giri National Labour Institute 2014); J.M.W.G. Lucassen, ‘Deep Monetisation in Eurasia in the Long Run’, in: R.J. van der Spek and B. van Leeuwen eds., Money, Currency and Crisis: In Search of Trust, 2000 BC to AD 2000. Routledge Explorations in Economic History (Abingdon: Routledge 2018). 2 Lucassen, ‘Deep Monetisation: the Case of the Netherlands’, 73, 87.
R.J. Stapel – Coin Evidence as a Proxy for Spatial and Temporal Changes in Wage Labour
ESSHC 2018
3
spatial and temporal spread of archaeological coin finds in medieval England and Wales to
reconstruct local spatial and temporal patterns of deep monetisation levels. I will focus on the
coins minted between 1180, when the English coinage was reformed and the short-cross
penny was introduced, and 1660. Finders of coins produced after that year are no longer
obliged to register their finds with the government. The proposed methodology can act as a
case study for other regions in the world.
Coin Finds and Production Figures
In order to reconstruct the spatial and temporal patterns of the level of monetisation in a
particular region, in this case medieval England and Wales, we need to combine a number of
datasets and variables. The basis for further analysis are the archaeological coin finds. Not all
coins that were in circulation can of course be recovered through archaeological surveys. The
metal of coins can be – and often is – reused for all kind of purposes, including the production
of new coins. Coins that were not reused or brought back into circulation, may have been
hoarded or lost. Of these coin hoards and coin losses, only a fraction will have been found –
and only a fraction of those will have been registered. We have to rely on the ability and
willingness of archaeologists and detectorists to find and record coin losses.
In countries that both allow the finder to keep his coin finds (after registration), and in
which a good system for registering the finds is in place, there are great possibilities for
researchers to use the coin finds evidence for socioeconomic and labour history. Such laws
and registration systems are for instance in place in the Netherlands (the NUMIS database),3
but the most advanced and user-friendly system at the moment is the Portable Antiquities
3 De Nederlandsche Bank, ‘NUMIS Muntvondsten Database’ (1997-2013) <http://www.dnb.nl/over-dnb/nationale-numismatische-collectie/numis/index.jsp> [accessed 2 May 2016].
R.J. Stapel – Coin Evidence as a Proxy for Spatial and Temporal Changes in Wage Labour
ESSHC 2018
4
Scheme (PAS) that registers archaeological finds – including coins – for England and Wales.4
Currently there are over 500,000 individual coins registered, mostly dating from the Iron Age
to the mid-seventeenth century5, as well as almost 1,000 coin hoards. On a daily average, 60
to 70 coins are added to the database.
Most of these coins have spatial coordinates attached to their records, but only
165,000 of these coordinates are publicly available. The spatial coordinates of the majority of
the records is concealed, to prohibit treasure hunters scanning the area for further artefacts.
In such cases, often only a parish or field name is recorded that can cover a couple of square
kilometres. For the purpose of this research, the publicly available data of the Portable
Antiquities Scheme was cleaned extensively and enhanced with an estimation of the spatial
coordinates – when absent. Wrong or missing delimiters were fixed, the spelling of
geographical names was harmonized, just like the names and estimated production dates of
the coins. The values of the denominations were harmonized wherever possible. In the end,
our adjusted dataset includes precise or approximate locations for 91% of all coin finds, a
significant increase from the publicly available coordinates.6 This means that there is a
substantial spatial coverage, that will make further analysis possible.
4 The British Museum, ‘Portable Antiquities Scheme’ (2013-2015) <http://finds.org.uk/> [accessed 2 May 2016]. 5 There is no obligation to report coin finds that were produced after 1660, so there is only a limited amount of modern coins included in the database. 6 All records that carried information on the parish or field or area name in which the coin was found, were given coordinates. Some coordinates are therefore not precise find locations, but center points of parishes. In those case in which only a district or county name is available, the spatial coordinates are only used for county or district aggregates.
R.J. Stapel – Coin Evidence as a Proxy for Spatial and Temporal Changes in Wage Labour
ESSHC 2018
5
Figure 1. Production of pennies (1 d.) between 1279–1351.7
Coin Name Denomination Survival rate
Farthing ¼ d. 0.00411% One in 24.330 coins
Halfpenny ½ d. 0.00536% One in 18.666 coins
Halfpenny (1335–51) ½ d. 0.00288% One in 34.721 coins
Penny 1 d. 0.00299% One in 33.400 coins
Gold noble8 20, 40, or 80 d. 0.00211% One in 47.334 coins
Table 1. Coins produced in England between 1279 and 1351, compared to the concerned coins reported in the Portable Antiquities Scheme.9
For the short period between 1279 and 1351 it is possible to get some indication of the survival
rate for each denomination of medieval English coinage (Figure 1 and Table 1). For this
7 Sources: M. Allen, Mints and Money in Medieval England (Cambridge 2014) 410–415; The British Museum, ‘PAS’, version: 17 March 2018. 8 This includes the – more frequently minted – half noble and quarter noble. Based on the coin finds of 1279–1351, on average out of each pound sterling, around 7.63 gold coins were minted. This includes 0.72 full nobles (9% – or 24% per lb.), 2.20 half nobles (29% – or 37% per lb.), and 4.71 quarter nobles (62% – or 39% per lb.). Since the production figures for all three types of gold nobles are aggregated and presented in pounds sterling, we had to transpose the coin finds to pounds first and then make the calculations. 9 The production figures are available in: Allen, Mints and Money, 410–415. Undefined and foreign coinage of this period that was recovered is not included.
0
200
400
600
800
1000
1200
1400
0
10,000,000
20,000,000
30,000,000
40,000,000
50,000,000
60,000,000
1279
1282
1285
1288
1291
1294
1297
1300
1303
1306
1309
1312
1315
1318
1321
1324
1327
1330
1333
1336
1339
1342
1345
1348
1351
Find
ings
of p
enni
es
Prod
uctio
n of
pen
nies
Produced Finds
R.J. Stapel – Coin Evidence as a Proxy for Spatial and Temporal Changes in Wage Labour
ESSHC 2018
6
purpose and because not all coin finds can be dated precisely, any date spans of the coins (e.g.
1307–27) are adjusted, where each year within this span is given a value of one divided by the
total span length: in this example each year gets a value of 1/21, or 0.0476. This allows us to
aggregate the year spans without producing duplicate counts.
It seems that gold coins are a little bit less likely to be retrieved than silver coins –
possibly because gold coins are more likely to be reused and less likely to be lost due to their
high value. One in 47.000 gold nobles, including half nobles and quarter nobles, is retrieved.
For pennies and farthings, roughly around one in every 20.000 to 35.000 coins produced is
reported in the Portable Antiquities Scheme database. The halfpennies show a slightly
different picture, with a particular high survival rate between 1292 and 1330. This probably
means that either the production figures for halfpennies are underreported, or the dates
reported in the PAS are incorrect. For the moment, a survival rate of one in 31.197 for small
silver coins (the average for all farthings, halfpennies, and pennies combined) seems
reasonable to us, especially as the survival rate does not vary that much throughout the period
1279–1351. Perhaps this figure can be further specified in the future, specifically regarding
whether this survival rate changes over longer periods of time (which is presumably a relevant
factor).
Length of Circulation Period
One of the other ways of interpreting this survival rate figure – which also adds temporal
depth – is by linking it to the length of the period a particular coin was in circulation. In order
to determine the per capita stock of coinage in circulation, we need to determine how long
coins remained in circulation since they were minted. In his study of the deep monetisation
levels in the Netherlands, Jan Lucassen used a ‘half-life’ of fifty years for copper coins –
R.J. Stapel – Coin Evidence as a Proxy for Spatial and Temporal Changes in Wage Labour
ESSHC 2018
7
meaning that as a ‘rule-of-thumb’ after fifty years, half of the coins produced were removed
from circulation due to losses, hoarding, melting, and reminting. For silver and gold coins,
which may have been more prone to hoarding and melting, this half-life may have been
shorter. Lucassen mentions examples of twenty to thirty years.10 From the composition of
medieval English coin hoards, it becomes clear that it is not unusual to have pennies included
that were up to 150 years old at the time the coin hoard was created.11 Using the survival rate
of thirteenth- and fourteenth-century coins we calculated earlier and a ‘half-life’ of fifty years,
we can draw a hypothetical survival rate below (Figure 2). The rate decreases exponentially,
with 1.5% per year.
Figure 2. Hypothetical survival rate for medieval small coins. The area in grey is used to calculate the estimated number of coins in circulation at any given point.
10 Lucassen, ‘Deep Monetisation: the Case of the Netherlands’, 90. 11 For farthings and halfpennies such calculations cannot be made due to a lack of evidence. M.C. Allen, ‘The Interpretation of Single-Finds of English Coins, 1279-1544’, British Numismatic Journal 75 (2005) 50–62.
0
0.2
0.4
0.6
0.8
1
0 1 0 0 2 0 0 3 0 0 4 0 0 5 0 0 6 0 0 7 0 0 8 0 0 9 0 0
SURV
IVAL
RAT
E
NUMBER OF YEARS SINCE PRODUCTION
R.J. Stapel – Coin Evidence as a Proxy for Spatial and Temporal Changes in Wage Labour
ESSHC 2018
8
Figure 3. Estimation of yearly small coin production (in pennies) per capita, based on the coin finds.
To apply this survival rate on the coin finds, we use the available coins dated in one particular
year (Figure 3) and combine these with all coins in the hundred years previous to that year.
This is done proportionally, using the expected loss of 1.5% per year (compare Figure 2).
Although after hundred years, still around 20% of the coins may be kept in circulation
according to the hypothetical survival rate above, the calculations are topped at hundred
years for simplicity. The end results are presented in Figure 4.
0
5
10
15
20
25
30
35
966
986
1006
1026
1046
1066
1086
1106
1126
1146
1166
1186
1206
1226
1246
1266
1286
1306
1326
1346
1366
1386
1406
1426
1446
1466
1486
1506
1526
1546
1566
1586
1606
1626
1646
Estim
ated
wor
th in
pen
nies
Coins worth an hourly wage or less (per capita) Coins worth a daily wage or less (per capita)
R.J. Stapel – Coin Evidence as a Proxy for Spatial and Temporal Changes in Wage Labour
ESSHC 2018
9
Figure 4. Estimated number of small coins in circulation (in pennies) per capita (topped at hundred years after production). The figures are based on the coin finds in the PAS database.
Population Figures
For the moment, I have used the national and county estimates for the population of England
prepared by Broadberry et al.12 These were interpolated to produce yearly estimates. The
population of Wales was added by adding 7% to the population of England.13
12 S.N. Broadberry, B.M.S. Campbell and B. van Leeuwen, English Medieval Population: Reconciling Time Series and Cross Sectional Evidence (2010), there tables 4 and 6 (national data), and table 8 (county data) <https://warwick.ac.uk/fac/soc/economics/staff/sbroadberry/wp/medievalpopulation7.pdf>. 13 This percentage is more or less stable throughout the years. Compare: G. Clark, ‘The long march of history: Farm wages, population, and economic growth, England 1209–1869’, Economic History Review 60 (2007) 97–135. I have use the data available at http://faculty.econ.ucdavis.edu/faculty/gclark/English%20Data/England%20NNI%20-%20Clark%20-%202015.xlsx; B.M.S. Campbell, ‘Benchmarking medieval economic development: England, Wales, Scotland, and Ireland, c.1290’, Economic History Review 61 (2008) 896–945; J.C. Russell, British Medieval Population (Albuquerque: University of New Mexico Press 1948) 360–361.
0
10
20
30
40
50
60
70
0
50
100
150
200
250
300
350
1066
1086
1106
1126
1146
1166
1186
1206
1226
1246
1266
1286
1306
1326
1346
1366
1386
1406
1426
1446
1466
1486
1506
1526
1546
1566
1586
1606
1626
1646
Num
ber o
f coi
ns w
orth
an
hour
ly w
age
or le
ss
Num
ber o
f coi
ns w
orth
a d
aily
wag
e or
less
Coins worth a daily wage or less in circulation Coins worth an hourly wage or less in circulation
R.J. Stapel – Coin Evidence as a Proxy for Spatial and Temporal Changes in Wage Labour
ESSHC 2018
10
Wage Levels and Hours Worked
Currently, I have used the nominal wage levels for skilled craft workers gathered by Gregory
Clark.14 Furthermore, for now, I have not taken into account changing patterns in the number
of work hours in a day. For the calculation, I presume there are eight work hours in a day, as
most wage series state the daily wage only. This also allows for good comparisons with the
deep monetisation levels created by Jan Lucassen for the Netherlands and several countries
in Europe and Asia.15
Due to the rising absolute wage levels in England from the thirteenth to the
seventeenth century, ever so often new coin denominations fall into the category of coins
worth an hourly wage or less. These coins are listed in Table 2.
The wage level between 1066 and 1209, the earliest year included by Gregory Clark, is
estimated to be static and the wage level in 1209 is used for all years. While this was most
certainly not the reality, it will not influence the deep monetisation levels too much, as
farthings, the only coin that matches the wage level of 1209, appear in this period in small
numbers – and in the form of a penny cut in quarters – only.
14 G. Clark, English prices and wages, 1209-1914 (Global Price and Income History Group, University of California – Davis 2006) <http://gpih.ucdavis.edu/files/England_1209-1914_(Clark).xls> [accessed 18 March 2018]. 15 Lucassen, ‘Deep Monetisation: the Case of the Netherlands’; Lucassen, ‘Deep Monetisation in Eurasia’.
R.J. Stapel – Coin Evidence as a Proxy for Spatial and Temporal Changes in Wage Labour
ESSHC 2018
11
Figure 5. Nominal daily wage of a skilled craft worker.
Period Largest available denomination equalling hourly wage
[1066]–1357 ¼ d. (Cut farthing); Farthing (introduced 1216–47)
1358–1559 ½ d. (Cut halfpenny); Halfpenny (introduced 1216–47)
1560–1616 1 d. Penny; Threefarthing (¾ d.; introduced 1561–82)
1617–1651 1 ½ d. Threehalfpence (introduced 1561–82)
1652–1660 2 d. Halfgroat or twopence
Table 2. Largest available denominations in circulation equalling an hourly wage.
Deep and medium monetisation levels: yearly estimates and spatial patterns
To calculate the deep (and medium) monetisation levels, the figures in Figure 4 are divided by
the typical nominal wage level of a skilled worker in England, either by hour (deep
monetisation) or by day (medium monetisation). The end result is presented in Figure 6. Quite
pronounced are the three shifts around 1350–55 (coinciding with the Black Death), 1550–60
0
2
4
6
8
10
12
14
16
18
20
1200 1225 1250 1275 1300 1325 1350 1375 1400 1425 1450 1475 1500 1525 1550 1575 1600 1625 1650
Penc
e
Wage, skilled craft worker
R.J. Stapel – Coin Evidence as a Proxy for Spatial and Temporal Changes in Wage Labour
ESSHC 2018
12
(mid-Tudor crisis), and 1650–55; each time followed by long but slowly decreasing levels of
deep monetisation. All three volatile periods are associated with either sudden or more slowly
developing rising nominal wage levels, causing new denominations to be included in the
calculations (see Figures 7 (deep) and 8 (medium)). Immediate population changes, affecting
the per capita numbers, play a role as well, particularly of course after the Black Death.
It is also possible to calculate regional, county estimates for the deep and medium
monetisation levels. Here we have to combine the county population estimates by Broadberry
et al. with the spatial data of the coin finds gathered by the Portable Antiquities Scheme or
added by my own. It builds on the presumption that coins minted often in London or
Canterbury end up circulating, either immediately after production or more slowly through
use and payments, in regions with the highest demand for that type of coin. The results are
presented in Figures 9 to 11.
The hypothesis that I wish to test is that regions with high rates of wage labour are
characterised by high demand of small denomination coins. If correct, this means that spatial
and temporal patterns in deep monetisation levels can be used to estimate the number of
people working for wages in a particular time, at least for rural areas. Indeed, areas we know
have had relative high wage labour rates in the Middle Ages, such as Norfolk, Suffolk, and
Licnolnshire, generally show high deep monetisation levels (Figure 10). The exercise is
particularly useful for the many regions of medieval England where such rates are completely
unclear.16
16 Some rural wage labour rates for England, including references to literature, are listed by Bas van Bavel: B.J.P. van Bavel, ‘Rural wage labour in the sixteenth-century Low Countries: an assessment of the importance and nature of wage labour in the countryside of Holland, Guelders and Flanders’, Continuity and Change 21 (2006) 37–72, there 38 (note 2).
R.J. Stapel – Coin Evidence as a Proxy for Spatial and Temporal Changes in Wage Labour
ESSHC 2018
13
In the coming year, I want to contextualise the outcomes of this paper, both in the
economic, monetary, and labour history of the British Isles, as well as adding comparisons to
other regions in the world. I will focus particularly on the reasons monetisation levels change
over time, and the possibilities of using these levels as a proxy for the presence of wage labour.
It might also be possible to help further define and/or sharpen the definition and premises of
‘deep’ and ‘medium’ monetisation levels.
Deep monetisation = ‘regular’ coin availability (valued one-hour wage skilled
worker or less) per capita. ‘Regular’ is approximated by
Jan Lucassen as 5 to 10 times the value of the prevailing
hourly wage.
Medium monetisation = ‘regular’ coin availability (valued full or half daily average
wage skilled worker) per capita
Figure 6. Monetisation levels in England and Wales, based on coin finds reported in the PAS database.
0
5
10
15
20
25
30
35
40
1066
1086
1106
1126
1146
1166
1186
1206
1226
1246
1266
1286
1306
1326
1346
1366
1386
1406
1426
1446
1466
1486
1506
1526
1546
1566
1586
1606
1626
1646
Deep monetisation level Medium monetisation level
R.J. Stapel – Coin Evidence as a Proxy for Spatial and Temporal Changes in Wage Labour
ESSHC 2018
14
Figure 7. Dissection of the deep monetisation level, per denomination.
Figure 8. Dissection of the medium monetisation levels per denomination.
0
5
10
15
20
2510
6610
8210
9811
1411
3011
4611
6211
7811
9412
1012
2612
4212
5812
7412
9013
0613
2213
3813
5413
7013
8614
0214
1814
3414
5014
6614
8214
9815
1415
3015
4615
6215
7815
9416
1016
2616
4216
58
Farthing Halfpenny Threefarthing
Penny Threehalfpence Halfgroat / Twopence
0
5
10
15
20
25
30
1066
1082
1098
1114
1130
1146
1162
1178
1194
1210
1226
1242
1258
1274
1290
1306
1322
1338
1354
1370
1386
1402
1418
1434
1450
1466
1482
1498
1514
1530
1546
1562
1578
1594
1610
1626
1642
1658
Farthing Halfpenny Threefarthing PennyThreehalfpence Halfgroat / Twopence Threepence Groat / FourpenceSixpence Shilling Twenty pence Two shilling
R.J. Stapel – Coin Evidence as a Proxy for Spatial and Temporal Changes in Wage Labour
ESSHC 2018
15
Figure 9. Deep monetisation levels (in absolute figures) for the years 1086, 1290, 1377, and 1600 (from the top left, clockwise).
R.J. Stapel – Coin Evidence as a Proxy for Spatial and Temporal Changes in Wage Labour
ESSHC 2018
16
Figure 10. Deep monetisation levels (relative to the other regions) for the years 1086, 1290, 1377, and 1600 (from the top left, clockwise).
R.J. Stapel – Coin Evidence as a Proxy for Spatial and Temporal Changes in Wage Labour
ESSHC 2018
17
Figure 11. Medium monetisation levels (relative to the other regions) for the years 1086, 1290, 1377, and 1600 (from the top left, clockwise).
R.J. Stapel – Coin Evidence as a Proxy for Spatial and Temporal Changes in Wage Labour
ESSHC 2018
18
Figure 12. Ratio of Medium to Deep Monetisation Levels for the years 1086, 1290, 1377, and 1600 (from the top left, clockwise).