A Growing Population in the Grip of Underdeveloped Agriculture

From the late 18th to the mid-19th centuries, Transylvania's population increased (according to the censuses of 1786, 1787, and 1850–51) by one-third, from 1.5 million to over 2 million, reaching 40 percent of the current density (see tables 1 and 2). The rate of growth, at around 0.43 to 0.47 percent per year, appears rather low. However, changes in the population of the border regions indicate that this level of increase was far from uniform. The rate of growth must have been lower until the 1820s, when, after the losses due to the Napoleonic wars, epidemics, and the 1817 famine, it registered a sizeable increase. In the 1830s and 1840s, bearing in mind what is known about the birth rate, the population grew at an estimated annual rate of 0.8–0.9 percent. Although contemporaries put the growth rate at one percent (which, if sustained, would double the population in seventy years), that figure was reached only in certain regions, notably the border districts and the Székelyföld. The population had probably never grown as rapidly as during these 'blessed times of peace.'[2]2. Report of the community of Szászivánfalva to the magistrate of Medgyes, OL, EOKL GP, Ügyiratok, 1847: 5563. Considering that in more recent times, the population of some so-called developing countries has grown at an annual rate of some 2–3 percent, the term 'demographic revolution' cannot be applied to Transylvania and the rest of Europe in the 19th century. Yet the rate in Transylvania was still remarkably high, for it approximated the rates that prevailed in England during the Industrial Revolution and, during a similar period of industrialization, in Lower Austria and Bohemia in the 1820s and 1830s.

{3-5.} Table 1: The number of population and the growth rate in Transylvania
on the basis of the population census in 1786 and 1850/1851

(on the basis of the disposition of lands in the years 1790–1848)*

Year
The counties and the Partium
Székelyföld
Szászföld
Borderland
Total
number
%
number
%
number
%
number
%
number
%
1785/86
991,674
168,591
134,144
1,575,130
1810
-
280,721
136,134
0.06
-
1821
-
146,726
0.68
-
1830
-
164,252
1.26
-
1840
-
179,908
0.91
-
1846
-
191,182
1.01
-
1850/51
1,245,000
268,000
187,000
-
2,073,202
1785/86-1850/51
0.35
0.72
0.44
0.51
0.43

* Data apart from those referring to Szászföld and the total number of population are estimations. After the events in 1848–49 public administration was reorganized, Székely frontier-guard regiments were disbanded. After the census Rumanian regiments were also disbanded.

{3-6.} Table 2: The density of population in 1786 and 1850/1851
(on the basis of the disposition of lands in the years 1790–1848)

Year
The counties and
the Partium
Székelyföld
Szászföld
Borderlands
Transsylvania
1786
25
15
41
24
25
1850/51
32
24
55
33
33
1850/51** The population of the borderlands was counted to that of the respective areas.
33
27
47
-
33

The present state of research does not allow for a full assessment of the extent to which social and regional variations in population growth affected ethnic distribution. Differences in the method of evaluation of unreliable data were so great that late-18th century estimates of the proportion of Romanians, who constituted an absolute majority, ranged from an unrealistically low 40–50 percent to as high as 60 percent; the proportion of Hungarians, who constituted close to one-third of the population, was estimated at 33–49 percent, and that of the Saxons at 10–18 percent (see table 3). By the 1830s and 1840s, the shifts in ethnic proportions that were brought by varying growth rates in individual regions largely balanced out on the national level, and the aggregate population's ethnic composition remained stable into the next century; and this despite the fact that flaws in the collection of data on ethnic origin introduced an element of distortion in the statistics (see tables 4–5).

{3-7.}

Table 3: The ethnical-national distribution of the Principality of Transylvania on the basis
of the contemporary denominational registers, estimations and the population census in 1850/51

Year
Rumanian
Hungarian
German
Gipsy
Jewis
Other
Total
Total number
of population
%
1766** We have two sets of data for this year.
58.9
27.5
13.6
-
-
-
100.0
953886
52.0
41.1
6.5
-
0.5
100.0
1453742
1773
63.5
24.2
12.3
-
-
-
100.0
1066017
1786
30.5
49.7
18.2
0.7
0.2
0.7
100.0
1664545
1794
50.0
33.0**
** 3–4 % Saxon or German.
12.5***
*** 3–4 % Hungarian.
0.1
0.1
0.1
100.0
2143310
1844
60.1
28.6
10.0
0.3
0.3
0.3
100.0
2143310
1850/51
59.5
25.9
9.4
0.6
0.6
0.6
100.0
2062379


{3-8.}

Table 4: The ethnical-national distribution of the Principality of Transylvania
on the basis of the population census in 1850–51 and 1930
(Territorial breakdown before 1848)

Ethnic group
The counties and
the Fogaras region
Székelyföld
Szászföld
Transylvania without
the Partium
1850-51
1930
1850-51
1930
1850-51
1930
1850-51
1930
Rumanian
781791
1203046
54246
102167
207810
320650
1043847
1625863
Hungarian
159396
319613
303975
440243
25063
68288
488434
828144
German
49166
56887
1163
2399
141425
177738
191754
237024
Jewis
10644
45229
1042
10370
165
9725
11851
65324
Gipsy
41117
41750
10022
11657
25244
16025
76383
69432
Other
6935
9638
2464
1724
1544
4492
10953
15854
Total
1049049
1676163
372912
568560
401251
596918
1823222
2841641

The demographic development of Transylvania is a distinctive variant of the east European pattern. One typical factor was the prevalence of early marriages, at around twenty years of age. In Transylvania, nearly 60 percent of marriages were contracted by couples in the lowest age category. In the same period, in Lower {3-9.} Austria, women under the age of twenty accounted for 6–8 percent, and men under twenty-four for 11–13 percent of all marriages; in Bohemia, the youngest age group of men and women accounted for one-fifth of marriages.

Table 5: Distribution of the three main Transylvanian nationalities
on the basis of the population census in 1850/51 and 1930 (%)

Area
Rumanian
Hungarian
German
1850-51
1930
1850-51
1930
1850-51
1930
Counties
74.5
71.8
15.2
19.0
4.7
3.3
Székelyföld
14.5
18.0
81.5
77.4
0.3
0.4
Szászföld
51.8
53.7
6.2
11.4
35.2
29.8
Transylvania
57.1
57.2
26.8
29.1
10.5
8.3

Marital age, however, is only one retrievable indicator of a demographic pattern. Since more and more exceptions can be found, population growth cannot be explained by the so-called demographic transition model, which has been commonly applied to eastern Europe on the eve of modernity, and generally to societies on the brink of industrialization, does not suffice to explain population growth, for there are too many exceptions. Indeed, it appears that, in many respects, Transylvania represents one of these exceptions. According to a well-known principle of demographic transition theory, the crude birth rate (births per 1,000 people) rises and the death rate declines, and the spread between these two indices widens year by year. Yet these patterns appeared only in certain eastern regions of Transylvania.

The aggregate statistics for Transylvania show that the crude birth rate and the death rate evolved symmetrically over some two decades, with a very slight downward trend. Yet normally, under {3-10.} conditions of industrialization, there is a sharp and parallel fall in both birth and death rates. In Transylvania, the crude birth rate fluctuated between 39 and 49 percent, and the death rate between 20 and 30 percent; similarly low levels were reached in the more developed parts the Habsburg empire only at the turn of the century, and in Hungary even later.

The demographic data was gathered by the Churches and collated by Transylvania's Gubernium before being forwarded to Vienna. Due to shortcomings in the registration procedures, the data is not fully reliable, and it requires much manipulation to produce a chart showing change over time. On the other hand, a similar pattern is revealed by the data for the border region's population (which was mixed in the villages with 'bourgeois' or 'provincialist' elements) and by the territorially concentrated Romanian border regiment of Naszód, for which far better records were kept; and also by the better demographic data for some individual villages. It was not accidental that in an otherwise comprehensive statistical handbook, published in Austria in the 1850s, Transylvania was omitted from a few sub-sections; clearly, the data from Transylvania did not fit the conclusion that the author, J. Hain, had drawn from the data pertaining to other parts of the empire, that birth rates and death rates declined as one moved from the south and east to the north and west.

To the extent that the demographic data is accurate, it presents the image of a relatively healthy population. In the 18th century, in France, only one out of two children lived beyond the age of twenty, and the average age was around twenty-five; in Transylvania, two out of three children lived past twenty and the average age was around thirty, or, by some estimates, as high as thirty-seven. Contemporary statisticians had always estimated the average age of Transylvania's population to be higher than that of populations in the more developed provinces of the Habsburg empire. These higher estimates may be attributed to remarkably low infant mortality rates, which are, however, among the more questionable contemporary {3-11.} statistics: whereas, in the empire's more developed provinces, infants accounted for over a third of all deaths, the proportion in Transylvania was reportedly under one-fifth.

Contemporary observers were less sceptical and attributed the low infant mortality rate to the healthful environment of the mountainous regions, for the 'survivors' seemed to be in good health. A French visitor to central and southeastern Europe was horrified to see the bare footprints of women frozen in the snow; he was equally struck by the fact that, during his week-long 'walk,' he saw no sick or disabled people, but encountered a great many elderly folk, some of them over a hundred years old. (This does not contradict the fact that, according to later and more reliable demographic statistics, the average age was higher in Hungary than in Transylvania.) It is not surprising if contemporary observers, impressed by such evidence, often overlooked the fact that the absence of epidemics and the efficacy of vaccination and other preventive measures contributed greatly to the excess of births over deaths. There were only two outbreaks of cholera, in 1830, when it claimed some three to four thousand victims, and in 1836, when it caused 14,000 deaths. In the latter year, cholera was responsible for one-third of the deaths in the border districts, and around one-fifth in the széks and counties. At the turn of the century, the chancellery ordered priests to charge with infanticide the parents of children who had not been vaccinated and died of smallpox; the measure appears to have had some positive effect. Benigni, a secretary at the military headquarters who was familiar with local conditions, did not mention vaccination in his 1816 work on the border region, but in the second edition, which appeared in 1834, he credited the increase in population to vaccination. In the 1840s, according to official reports, over half of the newborn were vaccinated. That did not stop contemporary observers from dwelling on the many deaths caused by smallpox, although these accounted for no more than 3–6 percent of the deaths. On the other hand, the general improvement in hygienic conditions in Europe may have contributed more than {3-12.} local factors to the lowering of the death rate. The region suffered no major epidemics, and the quarantine system on the eastern borders of the Habsburg empire was effective in keeping out devastating diseases such as plague. In fact, partly thanks to new quarantine controls along the Danube, population growth was not reduced by epidemics in the Romanian principalities either.

In the better-endowed regions proximate to Transylvania, the needs of a growing population could be met by expansion of the cultivated area. In Transylvania, despite the comparatively low density of population, the most severe economic problem of the 1830s and 1840s was the growing inability of agricultural production to keep pace with demand. By the end of the 1830s, the expansion of cultivation at the expense of animal husbandry had come to an end, and it is not likely that the area of cultivated land grew significantly between then and 1848. Significantly, over two decades, the Gubernium's auditors at Kolozsvár conveyed the same data on taxable land to the imperial statistics office in Vienna, making no effort to indicate minor changes. To be sure, such tax surveys have never been considered a reliable source of statistics, and more research is needed at the local level to assess the accuracy of data collection. For instance, a cadastral survey in the 1850s found that the area of arable land was 1 million cadastral acres (570,000 hectares) greater than had been estimated from tax assessments that encompassed noblemen's estates and were largely based on the proprietor's reports. The discrepancy might be partly explained by tax evasion, which was rampant (there were frequent reports of failure to report the full extent of landed property), and by inaccurate and unprofessional surveys. Yet tax assessments from the 1770s through the 1850s seem to reflect with some accuracy the changes under way, particularly with respect to growth in the area of cultivated land. To be sure, the purposely inaccurate reporting of landed property was common in the counties and Székely széks; as will be seen, this practice was an essential factor in 'social stability.' Data from Szászföld and the border regions can be regarded as {3-13.} more reliable, especially in the case of the latter: being exempt from taxation, the border guards had no interest in underestimating their property, and the military authorities were probably more conscientious in drafting their tax reports (see table 6).

The stagnation in agricultural production was well-illustrated in Benigni's accounts. In both the 1816 and the 1834 editions, he used the same estimate of grain output to calculate per capita production, which, in 1813, was 215 kilograms per capita, and, in 1830, some ten percent less. Between 1830 and 1847, the grain production per capita decreased in the border districts by another ten percent. The per-capita area of ploughland was approximately 0.46 hectare (0.81 cadastral acre) in 1821 and 0.36 hectare (0.64 cadastral acre) in 1846; meanwhile, the number of ploughs increased at the same rate as the population. The number of livestock was stagnating as well; the only exception was sheep, whose numbers grew after 1828. Sheep were all the more necessary since wool was the primary raw material for clothing. In the border districts, the number of sheep still surpassed that of the inhabitants. There, and in the rest of Transylvania, the plentiful supply of sheep depended on transhumance: between 50 and 66 percent of the sheep were taken to winter pasture beyond the Carpathians, in the Lower Danube region. (Similarly, the so-called purzsás shepherds grazed their considerably smaller flocks on Hungary's Great Plain, until the expansion of ploughland drove them back to Transylvania and the Romanian principalities, where the pattern eventually repeated itself.) From the mid-1800s onwards, the number of taxable sheep in Transylvania (excluding the flocks of the landed nobility) had steadily fallen, to around 200,000 in the 1840s; meanwhile, the number of sheep taken to pasture beyond the Carpathians was increasing and, in some years, surpassed one million. It can therefore be estimated that the number of sheep owned by landed noblemen, who were exempt from taxes prior to 1848, stood at around 300,000.

{3-14.}

Table 6: The extent and growth of the ploughlands of the taxable Transylvanian
inhabitants and the frontier guards on the basis of the assessment of taxes, 1772–1853

Year
The counties, the Partium
and Székelyföld
Szászföld
Borderlands
cadastral acres
index (1772=100%)
cadastral acres
index (1772=100%)
cadastral acres
index (1772=100%)
1772
391744
100
170989
100
100997
100
1791[1]1 - From 1791 data are converted to cadastral acres from the original “köböl” [férőjűség]. 2 köböls= 1 cadastral acre= 57120 square meters.
441571
11271
214471
125.42
-
-
1796
429665
109.68
215411
125.97
-
-
1804
-
-
-
-
122779
121.56
1812
431114
110.04
223162
132.51
-
-
1813
-
-
-
-
120493
119.30
1821
-
-
-
-
119377
118.19
1828
445757
113.78
239642
140.15
-
-
1830
-
-
-
-
123410
122.19
1847
-
-
-
-
123865
122.64
1848
450015
114.87
243336
142.31
-
-
1851[2]2 - In 1851 already on the basis of the general and proportional sharing of taxation.
-
-
1093288
-
-
-
1852-53[3]3 - On the basis of a cadastral survey and not on the basis of declaration as in the case of tax assesments
-
-
2163063
-
-
-

{3-15.} Given Transylvania's soil conditions, it took a decidedly good harvest to meet domestic needs for grain. According to contemporary estimates, when crop yields were only middling, at least one-fifth of the demand could be met only by the importation of grain from the Banat and Wallachia. Landowners from various parts of the country reported that the average fall harvest of wheat produced only three times the grain sown, although some put that multiple at five or ten.

In the last third of the 18th century, corn gained in popularity in Transylvania; it was more profitable per acre, and its cultivation and processing demanded less investment. At mid-century, one-third of the ploughland was planted with corn, 15 percent with wheat, 15 percent with rye, and 10 percent with a mixture of wheat and rye. Corn was also grown in household gardens, as well as on fallow land, thus reducing the area available for grazing. Although, in the long run, corn helped to sustain a growing population in southeastern Europe — as did potatoes in the northern parts of central Europe — it as regarded as a sign of poverty. Over a ten-year period, there would be one good corn crop, four average ones, and five bad crops, and corn came to symbolize backwardness, an obstacle to economic growth: 'When the yield is good, nobody wants corn, and when the yield is bad, no one has it.'[3]3. M. Engel, 'Észrevételek az úgynevezett Erdélyi Gazdaságról,' Nemzeti Társalkodó, 2/7 (1834).

Due to the uncertain yields, 'old-type' crises would still recur. When crops were bad, grain prices rose; the purchasing power of most people fell; and tradesmen and craftsmen, faced by shrinking demand, raised their prices to make up for the lower volume of sales. The slow price inflation of the 'quiet years' ended in just such a shock, in 1830, when grain prices doubled from one week to the next. This was perhaps the most severe crisis since 1817. However, Transylvanians were fortunate in that there had been no coincidence of cholera and crop failure. The cholera outbreak in 1836 actually coincided with a record crop. In 1833–34, and again at the end of the decade, famine — always a threat in the wake of a bad {3-16.} harvest — was averted by substantial imports of grain from the Romanian principalities and Hungary. Even tax arrears were reduced substantially until the mid-1840s, when the wheat harvest was so poor that half of the crop had to be replanted to 'safeguard the future.' When, in 1847, all of Europe was threatened by famine, Transylvania managed to escape that fate even though the harvest was below average; to be sure, in many places taxes could not be collected, but the only concrete measures taken to prevent an aggravation of the crisis included a prohibition of the dangerously spreading practice of distillation, and a limitation of the amount of grain that could be bought up by Hungarian merchants. Previously, the main response to crisis had been to import grain from Hungary. Now, however, the crisis had struck Hungary with greater force, and this had a severed impact on the people, mostly Romanians, in Transylvania's mountain regions, who relied on grain imports from the Banat and the Great Plain.

There is no record of crises as severe as those of the 1790s or the 1810s, but their memory lingered. In the 1830s, people were still apprehensive: 'A single barren year, a drought or a rainfall are enough to empty the pantries and drive the common people to earn their bread as itinerant labourers.'[4]4. K. Hodor, Doboka vármegye természeti és polgári esmértetése (Kolozsvár, 1837), p. 360. Itinerant workers and woodcarvers from the Érc Mountains had always relied on the grain earned by their labours; even in the 1850s, the people of Topánfalva would warn the authorities that 'a single crop failure in Hungary, coinciding with an average crop in Transylvania, could spell death for us,' much as it had in 1817.[5]5. Petition of the peasants and miners of Topánfalva, written by Nicolae Corcheş to the military governor, OL, EOKL, Militär- und Civil-gouvernement, Statthalterei in Siebenbürgen, Általános iratok, 1852, 115 p. 55. If bad weather reduced the corn crop, a million Romanians would 'lose their bread,' and thus, 'in our homeland, a year of high prices will not fail to drive a mass of proletarians to beggary.'[6]6. L. Kőváry, Erdélyország statisztikája (Kolozsvár, 1847), p. 161. The fear of crises turned out to be more enduring than the phenomenon itself. Stagnant output and productivity nurtured popular concern, threatening a social crisis, yet, in retrospect, the economic problem was more apparent than real.

However, Transylvanian society and its basic socio-economic unit, the family, had to find a response to these pressures. One preventive {3-17.} measure, and one that had become common practice in this period, was to choose the right time to found a family. Due to the rigid pattern of seasonal work and religious feasts, most marriages were concluded at year's end or at carnival time. Later, as in every agrarian society, there would be a rise in the number of marriages in years when there was good harvest of grain or grapes and when agricultural prices were high.

As noted, some sixty percent of Transylvanians married at a very early age. In some Romanian districts, thirty-year old grandfathers were not uncommon. Even Saxons favoured early marriage. Among Lutherans, 90 percent of whom were Saxon, over half of the marriages contracted in 1847 were between women under twenty and men under twenty-four. Among Roman Catholics, the proportion was closer to a third. The pattern varied by region and social stratum. According to contemporary accounts, prosperous peasants had fewer children then the poorer ones. By the beginning of the 19th century, the 'two-child system' (i.e., birth-control) was being adopted by more affluent Saxon peasants to compensate for early marriage. The new custom was intended to slow down the fragmentation of land through inheritance.

Notwithstanding the deficiencies of record-keeping, it can be inferred from the low (crude) birth rate and high (crude) marriage rate that birth control was widely practised. In an encyclical letter, the Greek Catholic vicar of the Naszód border regiment advised his priests that when they gave instruction about the Fifth Commandment, 'thou shalt not kill,' they should emphasise that women and maidens committed a grave sin when 'taking concoctions of diverse plants to prevent fertilization or to kill the foetus,' not to mention infanticide right after delivery.[7]7. Şt. Buzila, 'Documente bisericeşti,' Archiva Someşana (1936), p. 407.

A practice contrary to the general trend was gaining ground in some regions: the so-called Malthusian custom of delaying marriage. It was most prevalent in the Székelyföld, where it was reported that most nobles tried to forestall the fragmentation of their estates by marrying late. The general practice, particularly in the {3-18.} Szászföld, was for newlywed couples to live with parents until they acquired their own plot, but in the Székelyföld, noblemen preferred to delay marriage until they had won their share of the paternal inheritance. Thus the incidence of illegitimate births came to be attributed in no small part to late marriages.

In the Háromszék's border district, early marriage was common, yet (taking the average over eighteen years), one child in eighteen was born illegitimate; farther north, in Csík, where the population was ardently Catholic and lived in rural isolation, the ratio was only one in thirty-five. In the northern, border-guard villages of Naszód, where most people were Greek Catholic, the illegitimacy rate was one in twenty. Meanwhile, in the southern Carpathian region, the rate was only one in 158. This may be attributed both to very early marriages and to the customary requirement that the man marry the girl bearing his child; if he desisted, the illegitimate child would be adopted by the village community, as happened in the Érc Mountains [Érchegység].

Economic changes must have had an impact on family structure. The population had suffered major losses in the early years of the century, and while it began to grow again at a comparatively high rate, the number of tax-paying householders continued to decline between 1831 and 1847 (see table 7). Since only around fifty people emigrated each year, the most plausible explanation of this apparent contradiction is that the size of families was growing. If one sets the number of tax-paying householders against total (actually, 90%) population, then it appears that, between 1831 and 1847, the average family size grew by almost one. The number of householders had declined only slightly among the lesser nobility; the decrease was somewhat greater among burghers, and much greater among serfs and cotters.

{3-19.}

Table 7: The number and the percentage distribution of taxpayer families, 1805–1853

Type of family
1805
1821
1831
1847
number
%
number
number
number
%
Lower nobility** Owner of one hind of land, Armalist, Boer of Fogaras.
12529
3.63
10874
12721
12233
4.00
Tax-paying bourgeoisie
10752
3.12
14054
12947
10459
3.42
Free peasants and freemen**** Lófő (Székely nobleman), gyalogszékely (Székely commoner), Rumanian and Saxon from Szászföld, free inhabitant of a town.
81540
23.64
81965
89865
59468
31.26
Socage serfs and cottars
216295
62.70
174670
188645
165439
54.17
Miners
6780
1.97
5674
8764
9424
3.09
Gipsies
12353
3.58
9601
9497
9196
3.01
Jews
484
0.14
555
718
1459
0.48
Others****** Greek, Bulgarian, gold-washer, sailor, salt miner, etc.
4217
1.22
3771
1958
1753
0.57
Total
344950
100.00
301164
325115
305431
100.00

Only among freeholders and taxpayers in the Szászföld did the number of householders register an increase. Looking back from the vantage point of the 1860s, a journalist reflected that 'in those {3-20.} feudal days, it was customary to forego redistributing landed property during the father's lifetime. Even after his death, younger brothers would gather around the eldest to safeguard the integrity of the property; by living in communal fashion, they managed to maintain a decent standard of living, which would have been unattainable if they had subdivided the land. They lived by their labour, and fathers were scarcely distressed by a family expansion that brought more manpower: in poor families, each child represented a working asset.' Few of them went into domestic service: 'living on a large estate, they could always find some agricultural tasks and preserve their independence.'[8]8. Dániel Dózsa's editorial in Kolozsvári Közlöny, 25 October 1862, no. 151. This retrospective, and perhaps embellished image was painted at a time when such communal households had become less common. Poverty was probably a more decisive factor than indicated by the journalist. Foreign visitors were struck by the sight of 'several families cohabiting in a single miserable peasant hut that was bereft of flooring, ceiling, chimney, and partitions.'[9]9. C. D'Haussez, Alpes et Danube II (Brussels, 1837), p. 306. In fact, the statistical reports for Doboka County reveal that in the 1830s, the average family had eight to nine members. In the similarly poor Zaránd County, it was not uncommon that fifteen people shared a single-room dwelling; in 1847, some 2,000 families, with eight members on average, had to seek public assistance.

In the Szászföld, on the other hand, the average family size had been falling since the end of the eighteenth century, and it stood at four in the 1840s. In Nagysink szék, the average family size was not quite three; even the families of Gypsies, who were notoriously prolific, averaged only 3.5 members. Nevertheless, the number of families increased and thus, despite the 'two-child system,' holdings underwent such a rapid fragmentation that the Saxon Universitas decided in 1848 to set a minimum size, beyond which land could not be subdivided.

Pending further research, it can only be hypothesized that the cohabitation of several families was, in many cases, only a pretence {3-21.} designed to ease the tax burden. Many families would claim to be a single unit, yet maintain separate households, and cultivate inherited plots that had been redistributed in practice, if not in law. In the Székelyföld, necessity was such a hard taskmaster that it was 'not uncommon' to see 'five or six married sons and their children living on the plot that was worked by the father.'[10]10. E. Jakab, 'Telepítésügy,' Erdélyi Híradó, 10 June 1847, no. 255. This probably explains why the censuses registered more than seven persons per house and family.

Meanwhile, in border districts, people who had noble-freeholder peasant status, or who aspired to it, showed a preference for smaller families. In order to help the population bear the burden of soldiering, the military command tried to induce, by administrative measures, the border-guards to form extended-family households on the model that prevailed in similar districts of the Banat, Slavonia, and Croatia. People living at the address were classified as a single family even if they had separate dwellings and cultivated separate plots. The military authorities would not allow the separation of families or the division of landed property and tried to buttress the 'paterfamilias' role of the householder. On the other hand, the authorities considered that the population in border-guard districts was too small to allow the exemption of householders and eldest sons from military service, which was general practice in the empire; although, in the early 1800s, such people became exempt from state taxes, they found it a heavy burden to do military service for even the minimum of 140 days a year.

Despite the prevailing dire circumstances, the joint-household model did not always take root. There were examples of such harmonious extended families, but in the 1840s, when delegates from the villages of Naszód drafted a plan for improving their lot, they opposed this system. By conceding that householders lacked sufficient manpower for summer work, they tacitly acknowledged the economic advantage of extended families, but they chose to highlight the chronic friction among brothers and sisters, mothers-in-law {3-22.} and daughters-in-law, children and adults. It was probably such instances that inspired Andrei Mureşan, the Beszterce-born poet of the Romanian national awakening, to note in his handbook on child upbringing (published in July 1848) that 'due to old-fashioned pre-judice, children are condemned to suffer more than American slaves' or domestic animals.[11]11. A. Mureşan, Icoana creşterei rele (Braşov, 1848), p. viii. The delegates from Naszód also complained that the organization of communal tasks was impeded by the tensions attendant on cohabitation of large families, and by the fact that when the eldest son succeeded his father as 'paterfamilias,' he would prosper at the expense of the others. They demanded the reestablishment of the old 'Saxon economic system,' by which they meant small autonomous families. And while the military administrators remained committed to the system of joint households, more than one ranking officer acknowledged that the small, autonomous families were a better guarantee of efficient farming.

Thus, while average family size in the border-guard districts rose from six in 1810 to ten in 1847 (or, among the Székely hussars, whose equipment was more costly, from nine to sixteen), this can not be regarded as a natural trend. The pattern indicates not so much the voluntary adoption of an extended-family model based on communal labour as the growing economic difficulty of maintaining a small and autonomous family.