Carneros y Chuño: Price Levels in Nineteenth-Century Peru

Among other dark legacies, nineteenth-century Peru remains an example of statistical “dark ages.” The Peruvian state was essentially prestatistical; until recently, it lacked the need and capacity for systematic collection of economic, social, and demographic indicators on so nominally a governed people. The challenge of unearthing and preparing usable data thus falls to modern students of Peru.

In the past decade, historians have managed to create from scratch fruitful new estimates of Peruvian foreign trade and export activities like mining. They have revised figures on such elusive quantities as the expansion of the public sector and its internal and external debts. New statistics on demographics, land tenure, business and monetary history, tariffs, labor force, and gross national product have appeared.¹ Most of these advances still cover the fairly accessible external and formal sectors of Peru’s economy, however, and we are a long way from statistics worthy of formal economic analysis. Scattered archival sources do in fact exist for a variety of quantitative tasks. But a conspicuous bottleneck is our lack of a domestic price series—a critical tool for advanced analyses of even the meager statistics on hand.

This study offers new information on consumer goods prices from nineteenth-century Lima and attempts to work them into a standard cost-of-living index and long-run estimates of inflation. While preliminary, these price deflators may be used for measuring “real” economic trends, that is, what economists call growth at constant prices. Inflation, as we knew from such qualitative signs as popular bread protests, became chronic in republican Peru, and this fact has bedeviled historians’ efforts to accurately describe development. According to nominal expenditure data, for example, the Peruvian state would have expanded some ten times in the course of the guano age—surely an exaggeration if not controlled for the era’s inflation.² Moreover, even in their simplest forms, price data help illuminate the links between broad economic developments and everyday social life. In the nineteenth century, Peru slogged through two decades of late colonial economic turmoil (1800-24); two decades of postindependence depression (1824-45); and three decades of frenetic guano prosperity, before the export collapse of the mid-1870s. Prices are a clue to how these events struck ordinary Peruvians and to numerous other problems in social history.

Readers should note, however, that all measures of price levels are little more than an economist’s invention, and thus open to endless improvement and interpretation. Historical—even current—price data are notoriously poor for countries such as Peru, especially if culled from secondary accounts with ill-defined standards of collection. Moreover, most of the existing archival price series for Latin America come from the colonial era or from rural zones, and not from cosmopolitan centers such as the one purposely selected here. Understandably, then, we have as yet no comprehensive or methodical price indices for nineteenth-century Latin America.³ This exercise aims to spur further research, simply by showing its feasibility.

The opening section presents the new archival food price data used. The following section devises appropriate price weights for a nineteenth-century cost-of-living index. These prices and index then yield new estimates and periodizations of Peruvian price movements. Besides illustrating the steps to a historical price index, the study underscores the indispensability of social history evidence for quantitative tasks. And it concludes by exploring in detail some applications of the new statistics to the economic, social, and political history of Peru.

Food Price Data

Peruvian governments never amassed systematic price data in the nineteenth century. For fragmentary information, four published sources exist. In 1870, amid rising urban unrest over subsistence costs, a regional commission headed by Lima Mayor Manuel Pardo put out a 161-page report on inflation. The survey, consisting mainly of heated opinions on the roots of the problem, also contains wholesale data for 29 common meat and grain products, in order to compare prices in 1854/55 and 1869. Jorge Basadre presented some of these haphazardly collected prices, and Pablo Macera even tabulated them into a series, within a wide-ranging qualitative discussion of inflation at the height of the guano age.⁴ The commission’s data reveal a doubling of prices for some staples, such as lentils and chicken, over the 15 years.

More recently, the historian Alfonso Quiroz, seeking ways to assess the social impact of the “consolidation” of the internal debt, discovered new food price data for the period 1847-65. These retail prices were gathered from hospital ledgers at the charitable archive of the Beneficencia Pública de Lima, and cover 23 items of popular consumption, such as carnero (mutton), beans, chickpeas, rice, and chuño (Andean freeze-dried potatoes).⁵ The dramatic fluctuations evident in certain items led Quiroz to identify two peak “crises de subsistencia,” in 1850 and 1856.

Two other works lie at the extremes of this study, and some of their data can be incorporated. Marcel Haitin, after digging in accounts of hospitals and suppressed monasteries (in a host of archives), averaged prices for seven major food commodities of late colonial Lima (1799-1824), which reveal, he argues, a hidden dynamism in the regional agrarian economy. On the other end of the curve, following the grave monetary upheavals of the Pacific War, other Lima commissions investigated rising prices of some twelve major components of the Lima diet, which form an 1897-1906 series.⁶ Still other historians, such as Vincent Peloso, have explored qualitative aspects of Lima consumption, nutrition, and supply networks.

The archives are then the best and only source for new data. The appendix to this study presents my original archival evidence on 29 Lima subsistence goods from 1826 to 1873 (see  Appendix). These are averages of weekly findings; wherever evident, the prices are standardized into reales per pound (8 reales = 1 peso before 1862, 10 reales = 1 sol thereafter). Prices of minor products such as herbs, and price fragments, are omitted, but the appendix does include the usable (consistent) data from the published works just mentioned.

Some of the new prices, following Quiroz’s lead, are derived from bimonthly purchasing records of two hospitals, San Andrés and La Caridad, and of the Colegio de San Carlos, all maintained by the Beneficencia (for 1826, 1829, and 1840-65).⁷ These are retail prices, the favorite of price historians, and robust enough with more than a dozen readings a year, usually from their original market receipts. While the charity may have received some unusual bargains (on bread for instance), the market prices also were checked by a scrupulous comptroller. Similar institutional accounts for government-run convents were discovered in the Archivo General de la Nación (AGN): the infirmaries of San Francisco and Santo Domingo list excellent data for 1829, 1831, and 1833.⁸ Another mine of price quotations is the books of military hospitals (such as Santa Ana) run by the Ministry of War and Marine.⁹ Located as well in the AGN, they provide a complete series of food items for the 1830s, based on the typical inventory of purchase receipts. Unfortunately, archival records become sporadic after 1843; weaker series were encountered under the Ministry of Government for 1848-52 and 1859-65 (San Carlos, San Bartolomé, Santa Ana, and San Andrés).¹⁰

Except for the lacunas of 1843-46 and 1857-58, a complete and consistent archival price series now exists for 1830-65. Combined with other sources, a reliable and wide record for essentials exists from 1826 to 1873. When we include some of Haitin’s archival colonial prices (which lack vital mutton and beef), and interpolate stable years from fragments, the series runs from 1800 to virtually the Pacific War.

A most lamentable gap yet remains—the years 1874-79. The late 1870s were a time of severe monetary instability, a period whose data are critical for any price index that aims to connect our macrostatistics through the denouement of the guano age. These prices were fruitlessly sought in the array of available primary sources. Other historians will no doubt resolve the 1870s gap as they uncover purchasing records of nearby haciendas, monasteries, or military units. What cannot be resolved is a direct linking of price series from before and after the Pacific War.¹¹ The era’s chaotic devaluations and inflation (reaching perhaps 800 percent between 1879 and 1885) must sober any historian.

One notable strength of the prices is that they are virtually free of the seasonal variations that bedevil most preindustrial prices series. Lima lay in a semitropical agrarian hinterland, where crops and livestock could be supplied with great regularity. The diversity of foreign suppliers to the coast (of grains, beans, and other articles) also smoothed fluctuations in this consummate market economy. In fact, the only periodic disturbances perceptible at the archival level were the sharp man-made monthly price hikes that could occur during caudillo wars, when supply routes were blocked. (One sees this graphically, for example, at the climax of the civil war of 1834.¹²) A weakness of this series, on the other hand, is the dearth of data on liquors and other liquid refreshments—which played a substantial role in the daily life of Limeños. When in hospitals, convents, or school, such substances were substituted by a remarkable array of medicinal products and herbs, whose make-up may interest historians of medicine.

Other sources supplement the archival ones, but must be taken with caution. From time to time, newspapers and census takers published reports from city markets. Lima’s dailies sometimes started (and then stopped) regular and detailed notations of wholesale and retail prices. Among the finest data found were El Comercio’s biweekly “Plaza del Mercado” reports for 1864-65 and La Patria’s analogous columns for 1871-73. I also considered such samples as 1828 wholesale lists from El Telégrafo de Lima and 1837 CIF import prices (El Redactor Peruano) for the already significant foodstuffs trade.¹³ For 1860, there are (probable) wholesale plaza prices gathered by the indefatigable statistician Manuel A. Fuentes for his Guía histórica-descriptiva, administrativa, judicial y de domicilio de Lima. Like Fuentes’s other guides, this is a mine of untapped statistical material (useful here, for example, is his quantification of all agricultural goods brought into the city).¹⁴

British and U.S. consular dispatches often reveal price data, particularly for the grain trades that so interested both parties (after 1843, British consuls were systematically tracking flour and wheat prices on a world scale). One suggestive British memo from 1842 compares basic living costs in seven major Latin American capitals, no doubt as a ploy for higher salaries. Bear in mind that Lima, with its cramped agrarian hinterland, remained the dearest of all regional posts, confirming both the complaints of its inhabitants about high costs and the overvalued nature of its currency. Prices for food, housing, and servants were twice those of Santiago and Buenos Aires, even before the Peruvian inflation of the 1850s.¹⁵ Finally, along with such supplementary prices, I take Macera’s tabulated data for 1855 and 1869 (and his strong flour series) also into account, despite their sometimes uncertain provenance.¹⁶

Historically, Lima was the wheatbread capital of South America, making the price of bread a special problem. Most accounts report a constant, customary, or controlled price of bread throughout much of the century—the postura de pan available at bakeries at four to five “pieces” a real, or less for charities. In fact, the quality and weight of these loaves varied widely. Often with the connivance of the municipality, monopolistic bakers adjusted loaf size and adulterated flour grades to reflect their real costs. Therefore, where possible, wholesale flour prices constitute the bread series, and, where these are lacking, a correlated U.S. wholesale index is used (in the early republic, substantial North American and Chilean flour imports stabilized prices in Lima).¹⁷ The wobbly downward trend that results is more precise than the constant price of bread assumed in other works.

In Table I, all of the usable data are simplified into 1830-based indices for each of eleven major food categories. In this study, 1830 will be the reference year; by then, prices had clearly settled at traditional levels following the extraordinary supply disruptions of the independence wars. These indices are averages of the most consistent and reliable appendix prices, after their standardization into real/pound ratios from the maze of archival quantity units and, for some foods, differing quality grades. (The daunting challenge is safely converting such archaic and imprecise measures as botijas, cargas, costales, or sacos of rice.) The number of minor and occasional products is further reduced in the table, except insofar as a combined “Others” index represents prices of potatoes, chuño, chickpeas, maize, squashes, and soap.¹⁸

On the whole, despite the occasional obstacles placed by supply monopolies and municipal price controls, Lima’s populace clearly faced a diverse agrarian supply shed and competitive markets. Limeños were subject to short-and long-term price shifts in staples like meat, rice, beans, and cooking fats. Yet, even these disaggregate figures suggest trends. Most prices skyrocketed during the independence struggles, then quickly stabilized. By the early republican period, between 1830 and 1848, prices appear largely in decline, as northern chacras, haciendas, and trade continued their recovery from the wounds of war.¹⁹ Overall, meat prices descended some 20 to 40 percent, and similar falls affected rice, beans, noodles, sugar, and a host of minor goods. After 1848, the reverse set in: an inflationary movement. Rising prices appear more fitful and flagrant in some lines of consumption—poultry, beans, rice—than in others—sugar, pasta, and imported flour. (The late 1850s emergency decrees for duty-free food imports coincide closely in timing and products with the price hikes observed here.) By the 1860s, most prices would remain far above 1840s levels, and some, such as those for beef and beans, at more than twice their original norms. Many explanations, we will see, are available for these trends.

In spite of the clarity of the trends seen in Table I, individual prices frequently moved in contrary directions and degrees. In and of themselves, moreover, food data cannot reveal the overall impact of price fluctuations on popular welfare, much less be used to deflate nominal statistics and trends. Such tasks require the construction and application of a weighted cost-of-living index.

A Nineteenth-Century Cost-of-Living Index

A cost-of-living index, such as those based on modern survey analysis, is a weighted average of price relatives over time. Put simply, the statistic shows the proportional change in overall expenditure needed to maintain a fixed level of consumption, as measured in some typical “base-year” family market basket. If a later basket B, with identically weighted items, costs 15 percent more to purchase than the original basket A, then inflation is 15 percent.²⁰ While in some sense a culture-bound measure, a cost-of-living index provides a far more meaningful account of price movements than noting that certain living costs seem to wildly fluctuate. Economists often deploy more sophisticated statistics for macroeconomic GNP deflators, but the consumer-price index remains the most tangible and direct measure of all.

Peru’s government, of course, did not conduct consumer research in the nineteenth century. Appropriate expenditure weights must be devised from scratch, to build a cost-of-living index that can employ our new price data for inflation estimates. The earliest Peruvian attempt to scientifically gauge consumption patterns was in 1920; the first official cost-of-living index followed in 1925, though an incisive review by Shane Hunt calls its accuracy into question (among other defects, it showed bread costing over one-fifth of family income). In fact, the oldest reliable survey, if still problematic for some items, is the 1957 census of Lima/Callao family budgets. For comparison, these modern expenditure weights are displayed in Table II. In lieu of other methods, Hunt was forced to fit these weights to the published Lima 1855/69 price data to calculate deflators for his guano-age GNP figures.²¹ By his heroic guesswork, price levels rose some 75 percent between 1855 and 1869, or only 32 percent if one included a stable price of bread.

Historians, however, are bound to question the anachronistic application of modern weights to prices from the 1850s. In a century of change, large shifts were likely in relative prices, and thus in budget weights—as well as in tastes for such traditional delicacies as carnero and chuño. Other nineteenth-century inflation “guesstimates” use indirect methodologies without the aid of expenditure weights or even local price data. Such is the cliometric exercise of Carlos Boloña for the 1870s to 1900s, which imputes domestic price levels (an acute inflation) from divergencies between North Atlantic price indices and Peruvian rates of currency devaluation. This methodology cannot be applied to the earlier era, which lacks decent exchange rate data. Further, since it uses cross-national index comparisons, it is less reliable than the direct knowledge of price movements from a consumer index.²² The nineteenth century, for all these reasons, requires its own cost-of-living weights.

Constructing these historical weights is a difficult but not impossible task. The method followed here builds rudimentary expenditure weights from three sources: a model artisan diet from 1869, archival hospital budgets, and aggregate consumption expenditures for the city of Lima in 1837. The combined midcentury weights may then be compared to modern survey weights, though the aim is to create the most fitting index for the historical price record.

The closest find to a contemporary food budget survey is an exemplary artisan diet from Lima’s Escuela de Artes y Oficios. Fortunately, the 1869 regional price commission published this model, as their crude way of suggesting the impact of inflation on popular living standards. The artisans’ daily food intake was listed only in grams, at a total cost of 35.5 centavos.²³ With but minor headaches, unit food prices of 1869 can be multiplied against gram quantities to reveal relative food expenditures; these are reassembled in Table III as percentage weights of a full food budget.

The results are by and large compatible with food budgets analyzed from hospital archives throughout the century. We can say that the typical Lima family spent roughly 34 percent of its food budget on meats (we knew that Limeños were voracious meat eaters); 5 percent each on lard and sugars (and that their diet was unhealthy); and 8 percent on each of the staples of rice and beans. Furthermore, these proportions were homogeneous for the population at large, for it is known that significant class differences had not yet emerged in creole cuisine.²⁴

Two further adjustments were necessary. First, the undifferentiated meat weight—for carnero, beef, and poultry—is apportioned on a straight one-third basis, i.e., 11.4 percent each, since meats were substitutes for each other. Second, two of the original weights, for bread and fuel, are readjusted to realistic proportions of 21.0 and 10.8 percent respectively, as calculated from 1837 hospital outlays.²⁵ The need to get bread right is obvious for Lima. Fuel (chiefly firewood) is also a vital, if overlooked, item to include in a price index. Used in cooking and other household chores, it suffered sharp price vacillations as the century progressed (from rapid deforestation of the coast and the subsequent 1860s switch to imported coal).

Diets, however, only reveal the average weights of food outlays, and Limeños did not live by carnero alone. What proportion did subsistence play in overall family expenses? The best data can be derived from a remarkably detailed series of surveys of the city’s global consumption, such as the one produced by Fuentes for 1858. The optimal information, however, comes from José María Córdova y Urrutia’s Estadística histórica, geográfica, comercial e industrial for Lima province in 1837. This census forms a more complete estimate of area purchases; provides convincing evidence on the collector’s method; cross-checks with other sources (such as consular reports and hospital records); and avoids some bizarre relative price shifts of the guano age.²⁶

The census data in Table IV contain the necessary global expenditure weights: foodstuffs as a whole at 54.5 percent; clothing and home textiles at 11.5 percent; rents at 7.8 percent; and diverse goods and services at 26.2 percent. These weights represent general trends. They are, in fact, the averaged post facto consumption pattern of each and every Limeño. Only one adjustment was necessary to the original data; liquor and tobacco are considered as essential foodstuffs, something that the profligate inhabitants of Lima no doubt felt at the time.²⁷ Finally, the artisan food weights and urban expenditure weights are merged to yield the nineteenth-century cost-of-living index, the one ready for use in Table V.

How trustworthy are these independently derived weights? Apart from the checks already noted and literary observations of the day, comparisons with modern weights (Table II) reveal the expected trends and some insights into social history. A striking coincidence occurs in food expenses; nineteenth-century Limeños spent 54.5 percent of their incomes on food, and modern workers 55.6 percent. The two leading items, meat and bread, account for roughly half of all food consumption—though, as one would hope, less bread and more meat grace the tables of modern workers. Outlays on minor staples all fall within 1 percent of each other, but recent urbanites have modestly diversified their nutrition. Engel’s law is modestly in force.²⁸

Nineteenth-century consumers allotted 11.5 percent of their budgets to textiles, modern ones 10 percent; housing expenses rose from 7.8 to 12.7 percent. Both trends are predictable during industrialization in a burgeoning metropolis like Lima. Even diverse expenses appear closely matched, though this conceals some typical signs of urban “progress,” for example, rising transport and health costs. And Limeños still love, if with less abandon, to smoke and drink.

Besides verifying the comparability of the new weights, long-term comparisons may support an economic truism. Cultural influences, or an enduring complex of taste preferences, are a fundamental determinant in consumption patterns. A Limeño of the 1950s, despite all the changes in the city, would have recognized and shared the gustos of one of its inhabitants of the century before (more than those of a Japanese, Mexican, or Scottish consumer of either era). This stubborn tendency has been noticed by economic historians studying centuries of English price history, and it is recognized in the greater methodological caveats to cross-cultural than to historical price comparisons.²⁹ Thus, it is safe to use these coastal creole weights in Table V for all of the nineteenth century. But only at great risk could we extend them to the other Peru of the Indian sierra.

Measuring Inflation

An actual price index, built from price information and budget weights like those above, is still at best an educated guess at changes in price levels. The statistician can choose to manipulate weights and data to achieve a lowerbound, upperbound, or realistic aggregate of the prices at hand; and in this mathematical task, too, the historian must integrate social history evidence. Out of the range of possible calculations, we now construct two distinct price indices for 1800 to 1873: first, a simple retail food price index, and second, a more elaborate and more weighted index of general inflationary trends. Depending on its intended use, the optimal measure will be the most cautious, comparable, and verisimilar one.

The first index simply measures the internally weighted aggregate shift in food prices. As a conception of inflation, it assumes that agrarian prices determined the price level in general—the not-uncommon physiocratic notion that all prices rose in tandem with the fortunes of agriculture. The starting year (100) selected is 1830, although it is not technically a “base year,” since our weights are an amalgam from the years 1837/69. By 1830, prices had stabilized after the inflationary crisis of the independence wars; it was a year of robust data and the base year of other series needing deflators.³⁰ By itself, a subsistence index is appropriate to certain tasks in social history, for example, understanding the motives behind the bread and bean riots so common in nineteenth-century Lima. In this case, its strength lies in the particularly solid set of local food prices found.

The calculation is a standard Laspeyres index. First, the food expenditure weights (set at 100, as in Table V) are multiplied against the 1830-based indices of itemized price rises (Table I). To be precise, the index is a weighted measure of changes in price levels relative to 1830. So, to illustrate, an indexed number of 88 (as seen with lard in 1829) represents a 12 percent decrease in price; when multiplied by its weight of 4.9, this produces a component change of –60. Alternatively, the index number for chicken in 1850 (109) represents a price rise of 9 percent since 1830, which after weighting by 11.4 contributes a score of + 103.

More than five hundred such calculations make the building blocks for the final index. The second step, then, is to add each year’s eleven food item total. When deflated by one hundred, this number reveals the year’s aggregate change in price level. (For example, the total of 962 in 1839 = +9.6 = the index number 109.6, or 9.6 percent higher prices than in 1830.) Along the way, we can prudently resolve the last gaps in the series.³¹

The result, in Table VI, is metropolitan Lima’s retail food price index from 1800 to 1873. Recall that this is our most promiscuous measure of deflation/inflation, one that assumes that all prices reflect the volatile behavior of foodstuffs. Yet it is still more accurate than impressions gathered from individual food fluctuations, for it is weighted, and these weights include bread and other price-stable items.

The food index demonstrates a late colonial deflationary trend (10 percent), interrupted by a 51 percent independence-war hike peaking in 1822. A rapid stabilization followed, with a continuing deflation in the 1830s, despite two caudillo-war crests. By the mid-1840s, aggregate food prices had descended 10 to 12 percent below 1830 levels, or more than one-fourth (28.7 percent) since the start of the century. As early as 1849, a fitful guano-age inflation began. Food prices rose 6 percent between 1848 and 1853, followed by significant spurts in 1854-56 (36 percent), 1859-60 (35 percent), and 1863-66 (33 percent). Overall in the high guano age, food prices had doubled.³² By contemporary norms, felt in the pockets of workers, this was a startling inflation.

Of course, foodstuffs were not the sole expense for consumers. A more cautious approach would further weight this food index in the overall consumption basket, assuming that other prices (for clothing, for example) remained stable. This is also shown, at intervals, in Table VI. (Calculation is simple: the food price index is deflated by its proportion in budgets, .545 × the food index; for each year .455 × 100 is added, which has the effect of holding all other prices constant.) Even if conservative, the assumption of price stability in other goods is not unfounded, given the era’s downward pressures on tradables; nor was it ever realistic to assume that, in a city, prices reflected agrarian conditions alone. When weighted this way, all price movements appear roughly half as strong. Prices fell only 11 percent from 1800 to 1830, and the total inflation from 1830 to 1873 was only 27 percent.

On the other hand, an annual food price index is not always sensitive to the wild fluctuations that could occur in moments of crisis. During the independence struggles, for example, blockades, hoarding, and spiking spread panic among Lima’s vulnerable population. To illustrate, we can use a bakery price series unearthed by one historian for the war years 1812-21 (during which bulk wheat prices swung between 3.5 and 25 pesos!). When factored into our index, this suggests that food prices may have doubled at the height of shortages.³³

The second major index to calculate is a measure of general inflation. In this, we integrate the impact of falling prices of imported manufactures, rising urban rents, and the overall effects of inflationary behavior and expectations. This index assumes that all prices shifted together— as a weighted average of the prices of foodstuffs, imported textiles, and rents. The most realistic and warranted model of prices in Peru’s opening economy, it is the one most likely followed by consumers, workers, and entrepreneurs, and it is congruent with many descriptions of the age. As the merchant J. F. Lembecke testified in 1869, “everything has gone up in proportion—wages, crafts, salaries, rents. . ..”³⁴

Separate indices for textiles and rents are found in Table VII. By 1830, North Atlantic textiles had flooded traditionally sheltered Peruvian markets, although their effects (through contraband) were perceptible as far back as the eighteenth century. Chiefly cheap industrial cottons, textiles remained the bulk of Peru’s nineteenth-century import bill (70 to 90 percent by some accounts); by 1840, they dominated the clothing sector in Lima, and they set the price of cloth in virtually all parts of the country.³⁵ In a sense, cottons serve as proxy for all foreign manufactures, though their prices are the ones that fell most sharply with the English industrial revolution. It should be kept in mind that the initial expenditure weight for clothing and home textiles was 11.5 percent.³⁶

The textile index is constructed from two distinct sources, for greater realism. For 1800 to 1830, I have used prices of U.S. sheet cottons. Until 1830, New England shippers were actually the major suppliers of “tocuyos” to Peru. These prices are also the best indication of what Peruvians would have experienced at the far end of a market interrupted by international wars and various obstacles to trade, including colonialism and initial republican tariffs. The index, like other sources, reveals an unstable trend to 1820, followed by a 60 percent drop in prices during the 1820s.³⁷ After 1830, when the English came to dominate the market, I use British FOB prices of export cottons. These underscore the delayed but full impact of the industrial and transport revolutions on Peru’s opening economy, with its progressively falling tariffs. By the 1840s and 50s, cloth prices had declined another 50 percent since 1830—a considerable savings for consumers—but they did rise again in the 1860s with the “cotton famine” and other inflationary pressures of the era.³⁸

Housing prices also varied dramatically in the nineteenth century. In fact, by the 1870s, rent hikes became a matter of grave official and popular concern, on a par with food issues. After actually receding during the postindependence depression, Lima’s population exploded during the guano age, largely through internal migration in the 1850s. The city doubled in size from about 55,000 to 120,000 by the 1870s, and rents soared accordingly.³⁹ Another influence was the concentration of Peru’s plutocracy in the capital, where the newly rich frenetically bid up housing prices.

Our best estimate of housing prices comes from the 1869 price commission. Average rent increases were placed at 8 to 10 percent annually between 1855 and 1869 (shop rents inflated even faster at a 12 to 15 percent rate). If this seems outrageously high, urban real estate taxes (those on predios urbanos) lend corroborating evidence: their sum tripled from 1850 to 1870, and then doubled in the early 1870s.⁴⁰ In Table VII, an index at 9 percent is calculated with base year 1854. Both the index and the tax data suggest a quadrupling of rents by 1873. Such a wrenching relative price change might render the rent weight (7.8 percent of budgets) rather conservative, since it reflected conditions in 1837. At the same time, this housing inflation can implicitly capture changes across a wider range of domestic services.

Before 1855, to be sure, housing prices moved along with other prices in the regional economy. They were stationary or falling until 1845, with the economic recession and population loss in Lima. This trend is clear from the tax records and from declining area farm rents. Rents likely rebounded modestly with the demographic and business revival between 1845 and 1855, much like other prices, before skyrocketing thereafter.⁴¹ The way to represent this pattern is simply to omit housing from the index until 1855.

The general inflation index in Table VIII thus combines two calculations. It is a linked index. First, for 1800 to 1854, the original food and clothing weights become 100, or the total change in prices (54.5 + 11.5 = 66 is recast as 100 = 82.6 + 17.4). The two reweighted series are then summed for each year. Second, for 1855 to 1873, the multipliers for food, clothing, and housing become, in similar fashion, 73.8, 15.6, and 10.6 respectively, and the three components are added for those years. Both calculations assume that all prices (even those omitted) behaved in tandem with the large portion of prices accounted for in the weights. The indices link in 1854-55.

Linking these two distinctly weighted indices is sound, with their minimal divergence in the years 1854-55. Moreover, it is methodologically desirable. For in this way we actually do capture part of the long-term relative price change that must have ensued when consumers shifted their midcentury savings from cheaper textiles into their higher rents.

Table VIII also contains a specific guano-age index, intended for statistical work on the post-1850 export era. The model and weights are identical to the second set in the general inflation index, which was multiplied by 1.202 to make 1854, rather than 1830, “100.” In 1854 prices began their ascent, after more or less stationary behavior during the initial decade of the guano age. This index graphically isolates the guano-age change in price levels: 30 to 60 percent by the first decade after 1854, and 70 to 100 percent overall by the following decade.⁴² If it was a dramatic shift, the explanation lies mainly in the mid-1850s trough in meat, flour, and textile prices. Consumers lost these benefits altogether by the 1860s. But that is precisely how inflation must have felt—alarming—to the generation living through the guano age.

Of all the indices, the long-range 1830 measure of general inflation is the most cautious, comprehensive, realistic, and solidly grounded. It represents, in my opinion, the optimal cost-of-living indicator, and can also be used as a statistical price deflator.⁴³ Peruvian price levels had clearly marked stages in the nineteenth century. Price levels were dropping slightly in the final decades of colonial rule, even if interrupted by the inflationary spiral of the independence wars. Two decades of deflation followed, bottoming out at 12 to 18 percent below 1830 levels, as the guano age began. In the mid-1850s, rapid inflation struck, which continued (with sharp fluctuations) throughout the remainder of the export era. By 1873, the price level stood 73 percent higher than 1830, or 32 percent above colonial norms. Beyond that point, Peru’s mercurial price history remains to be explored.

Periodization and Origins

By now, the strength and limits of these data and methods should be readily apparent. Price series remain ever open to additions and improvements, but these numbers do offer more precision and meaning than the qualitative guesses of the past. Through Figure 1, which portrays the food price index and the inflation index, readers may follow changing Lima price levels from 1800 to 1873. This supports a new periodization of price history—and may also help sort out possible factors behind the movements of price.

From the perspective of aggregate prices, the new periodization looks like the following:

1) 1800-14: Moderate Deflation. Possible gains in agrarian productivity, or more efficient commercial practices, were already at work on the ancien régime, according to these thin data. However, the Bourbon shipping and political crises of the era offset much of the downward effect, as Peru could not fully benefit from the price falls of the infant industrial revolution. This incipient deflation contrasts with inflation surmised from unweighted agricultural series, or as seen in agrarian prices in late viceregal Cuzco, Bolivia, and Mexico.⁴⁴

2) 1815-24: Sharp War Inflation. The supply disruptions of the independence struggles are notorious—the blockades, urban shortages, exactions of marauding armies, devastation of valleys such as Chancay and Mantaro, the flight of slaves. Together, they pushed up prices nearly 40 percent by 1822. This remains a timid estimate; transitory food shortages alone may have shot prices up twofold. The pressures to declare “free trade” in 1821 thus become even clearer. Monetary influences are harder to decipher, for San Martín’s unbacked fiscal emissions of 1822-23 were likely countered by the era’s deflationary losses of metallic money supply. This capital-bullion flight reached some 27 million pesos between 1819 and 1825. Relative price changes are also hard to pinpoint: prices of imported manufactures plummeted after 1821, just as those of domestic foodstuffs soared. But the impact appears erratic until the import “oversupply” crisis of the mid-1820s.⁴⁵

3) 1825-46: Stabilization and Deflation. Price levels stabilized remarkably well in the aftermath of independence, continuing their slide until the mid-1840s. Two caudillo-war peaks alone disrupt this trend. Indeed, once allowance is made for all three military crises (1815-24, 1833, and 1836-39), a secular deflationary movement is visible ever since 1800. Prices were sinking about 1 percent annually. This deflation has not been recognized before, except in the slump afflicting rural estates.

Several supply factors likely account for the 18 percent deflation from 1830 to 1846 (or 27 percent from 1826). The price falls involved both domestic foodstuffs and tradables. Until 1845, political-military instability kept Peru in deep economic recession, yet estates were able to quickly reconstitute. However, their only market was a depressed one. Most coastal farmers survived by switching from Pacific exports (sugar, tobacco) to provisions for local consumers, at a time of shrinking town populations and demand. On the tradables side, textile prices took their sharpest-ever plunge in this period (more than 50 percent). This time consumers perceived the savings, as they combined with Atlantic shipping advances and the dismantling of Peruvian tariffs (completed in textiles by 1840).⁴⁶

Local and international monetary factors were also at play. Peru’s chronic balance-of-payments gaps of the period—resolved by continuing exports of coin—may have permitted these downward macroeconomic price adjustments. By the 1840s, the revival of silver mining and the diversification of exports would have halted this effect. In any case, the impetus to deflation in Peru must have been formidable. Most contemporary analysts, in contrast, had predicted inflationary pressures from currency debasement (the influx from mid-1830s onward of “feeble” Bolivian coin), or had envisioned elevated food prices from the era’s stiff agrarian protectionism.⁴⁷

4) 1846-54: Moderate Guano Reflation. By the mid-1850s, prices were beginning to wobble up toward 1830 levels. The notable finding, however, is that the early expansionary effects of social peace and guano exports did not translate rapidly into an inflation. While urban population, business, and demand rebounded quickly in the late 1840s, this did not, for example, spawn supply bottlenecks. Domestic food prices did not leap until the labor crisis of the 1854-55 civil war.

Overall, this reflation does indicate spreading domestic demand felt from guano exports—contrary to enclave models of the bonanza, which argue that the export produced negligible income effects in the domestic economy.⁴⁸ Latent inflationary pressures may have been developing; yet they were contained by the continuing descent of import prices (in wage goods like grains and textiles), which was readily absorbed after Peru’s Manchesterian free-trade law of 1852.

Quiroz has specifically analyzed monetary factors in this period. He rejects the contemporary polemic that the massive internal debt consolidation of these years (nearly 24 million pesos by 1853) fueled inflation; the disbursed public debt bonds did not function as a “quasi-money.” The timing and degree of inflation shown here fully support that argument.⁴⁹ However, one should not discount excess demand as a distinct result of this redistribution, even if part of the spending leaked abroad by way of luxury imports.

5) 1855-73: Severe Guano-Age Inflation. Price levels rose precipitously in the mature guano age: some 70 percent by 1865, and 108 percent overall by 1873. As contemporaries noted, the civil war of 1854-55 marked the start of acute inflation; sharp fluctuations also struck in 1859-60 (during a major export crisis) and in 1864-66 (during the costly war with Spain). By the mid-1860s, however, these events clearly formed part of a secular inflation.

The years 1854-56 represent a conjunctural crisis in food supply (whose prices rose 36 percent) which led to emergency decrees for duty-free food imports in 1855, 1857, and 1859. But freer trade could not abate an encompassing inflation, now felt on every front—in domestic housing costs and, if less dramatically, in consumer tradables (textiles, grains) and domestic agriculture. By the 1870s, this pattern of inflation had completed Peru’s long-term shift in relative prices, the one first glimpsed in the 1820s: a cheaper tradables sector was edging out domestic production. The causes of inflation must have been the broadest ones, such as the rapid loan-financed growth of the state; the dramatic expansion of Lima’s population; and the proliferation of local banks issuing unbacked paper notes.

Since, by the 1860s, inflation had become an overtly political and popular controversy, homespun theories arose to explain it. For the outbreak of the crisis in 1854-56, historians discuss short-term stresses: the dislocations of war, the impact of slave manumission on coastal agriculture, and, with least evidence, the effects of Andean epidemics. The 1869 regional price commission,⁵⁰ by then examining a chronic problem, spoke of coin depreciation (the impact of monetary conversions in 1857 and 1863); the 1864-65 conflict (more clearly reflected by this index); the general “expansion” of Lima (i.e., demand factors); and persisting agrarian labor shortages. On weaker grounds, contemporaries and historians alike fault the transformation of nearby farms into export-oriented cotton and sugar plantations, without considering the probable efficiency gains of increasing food imports.⁵¹

Demography, supply conditions, and wages exhibited peculiar and revealing patterns during this period: perhaps a structural basis for urban inflation. The 1850s influx of rural and foreign migrants to Lima (that doubled urban population) spurred housing inflation, as did the concentration of Peru’s emerging plutocracy in the capital. For the first time, urbanization also strained local food supply sheds, as seen, for example, in the surge in meat prices and fuel prices (from sheer deforestation). All the while, Peru’s intensifying rural labor shortage helped maintain higher food prices. The mystery is why this inflationary social pattern persisted for more than two decades. Falling (real) urban wages and rising coastal agricultural wages should have reversed such demographic pressures on food and housing supplies.⁵²

Still, inflation could not have flourished if not allowed by expansionary monetary conditions—just as Peru’s previous deflation had occurred amid a shrinking money supply. The post-1860 growth in Peruvian money supply was caused by the sudden birth and multiplication of a Lima banking network (linked to the guano trade) and by the influx of foreign credit (linked to expansion of the state). In the 1870s alone, fiscal notes increased fourfold.⁵³ The abundance of exchange was reflected in the rapid fall in interest rates, approaching negative real rates, even as inflation raged on. Moreover, until the mid-1870s export crisis, none of these expansionary forces were constrained by demand-soaking tools such as monetary controls, devaluation, taxation, or recession.

6) 1874-1885: Inflationary Chaos. . . to Liberal Control? Direct data on prices are lacking for the last quarter of the nineteenth century, but other evidence suggests some trends. Inflation escalated during the crisis-ridden late 1870s, according to exchange rate analysis and contemporary accounts. A monetary inflation seems probable, even if attenuated by silver-standard depreciation (starting in 1872) and the economy’s export-credit collapse of 1875. Intense monetary instability accompanied the Pacific War and its aftermath (1879-85); price levels may have risen 800 percent.⁵⁴

By the 1890s, however, prices had stabilized, and Peru’s modern record begins. Policy makers may also have absorbed at least one painful lesson from the nineteenth century; until the present nightmares, an orthodox liberal Peru managed to contain inflation better than most countries of the region.⁵⁵

Even new price data are unlikely to alter these broad stages of price movements, except for the 1870s. But understanding the causes and impact of inflation is likely to advance well beyond this unavoidably sketchy and eclectic analysis.

Applications and Analysis

Sparse data, price theory, and common sense all dictate the cautious use of a historical price series. Table IX consists of deflators, based on the inflation index, for use in further statistical work. What, then, are some of the pressing practical applications of price data to the economic, social, political, and regional history of the Peruvian “dark ages”?

1) Economic Analysis

Price movements can help illuminate murky domestic economic developments during Peru’s first century of independence. Usually, these developments are glimpsed only through external indicators, such as export performance. Caution is needed in such analyses, however, since, to take one example, falling prices could reflect either economic contraction or growth-enhancing productivity changes.

For the 1825-45 period, Peru’s deflation is another sign of the caudillo-era depression, and prices can help us understand its sources. For example, how well was the international price/specie flow mechanism operating in Peru’s risky and disarticulated economy? According to classical theory, Peru’s domestic price structure should have leveled in response to early trade gaps and bullion-coin exports—i.e., a decreasing local money supply. Tradables (imports) should have become dearer relative to home goods, ending this typical disequilibrium of the nascent Latin American republics.

But, did the modest domestic price declines actually observed (no more than 12 percent) represent an efficient or inefficient local adjustment process? Relative to plummeting foreign prices (more than 50 percent in the 1820s-30s), they appear to have lagged. A faulty response, caused by any number of institutional or social factors, would have prolonged Peru’s depression, by forcing quantity/output contractions instead of adjustments of price.

These issues are particularly relevant because Peru, like most of the region, was struggling with the first shocks of its nineteenth-century transformation to an open economy.⁵⁶ Why did Lima’s economy always appear overvalued relative to foreign currencies? This could have been a factor in the economic and social dislocations that plagued Peru’s early attempt at trade liberalization. Specifically, did sticky price behavior add fuel to Peru’s initial (neomercantilist) protectionism, aggravate unemployment, or lead to a faster dominance of foreign goods?

Domestic price levels, not only overseas demand or the terms of trade, might also have affected Peru’s export performance. New studies, for example, show that Peru’s demolished silver mines of the Cerro de Pasco district actually recuperated quite rapidly in the postindependence period. By the early 1840s, output nearly reached colonial peaks. Deflation could have been one production incentive at play, by raising the real local purchasing power of silver. Conversely, Peru’s mining sector quickly and mysteriously collapsed again at midcentury; inflation might have become one adverse cost-price factor.⁵⁷ This pattern of rise and decline was followed by most of Peru’s other exports—apart from guano, which, among its other peculiarities, was peculiarly insensitive to domestic costs.

After 1845, the guano-age recovery is discernible in Lima’s halting reflation, and by 1855, in the inflationary spiral unleashed by rising domestic demand. Guano drove up incomes and spending in the domestic economy, at least around Lima, where the benefits concentrated. On the other hand, existing estimates of export growth should be moderated in real terms. Overall, Hunt figures that guano sales brought a total of 454 million pesos to the Peruvian economy—for an extraordinarily high 71 percent “returned value” ratio.⁵⁸ (This, plus the evidence of inflation itself, refutes the traditional enclave interpretation of the guano boom, i.e., the thesis that most guano income directly bypassed the local economy as payments to overseas interests.)

However, despite Peru’s favorable terms of trade, by the 1860s inflation was eroding the real domestic value of the bonanza. How much guano benefits fell would have to be computed on a yearly cumulative basis, but Peruvians themselves readily sensed this dilemma at the time. To many, inflation was the most palpable sign of the economy’s distress. It is one reason (besides fears of export exhaustion and mounting foreign debt) that this enormously prosperous period always seemed enveloped in an atmosphere of impending crisis.⁵⁹

An earlier study of the Lima economy provides an example of the need to measure, with a price index, the real growth in the domestic economy. It was a detailed time series for Lima business revenues, derived from archival patentes tax registers. The aim was to pinpoint the differential effects of export trends on local commercial, manufacturing, and service groups. In nominal terms, the aggregate index ran: 1830, 100; 1834, 91.6; 1839, 91.8; 1844, 71.3; 1850, 98.3; 1857, 139.6; and 1861, 155.4. These numbers suggested a severe business slump by the 1840s, followed by a dramatic recovery at the outset of the guano age. However, when deflated by the new price index, the series (in constant 1830 pesos) becomes less dramatic: 100; 94.7; 89.3; 86.7; 119.7; 128.8; and 127.7 in 1861. Over three decades, business expanded by 28, not 55, percent. This draws a more realistic portrait of the extent of both depression and growth in republican Lima.⁶⁰

The most challenging problem of the guano economy is also its largest one: did Peru experience any substantial and enduring productivity gains during this export bonanza? Even Hunt, who created a GNP estimate for 1876-77, remains in doubt. The clearest test, analogous to work on nineteenth-century Mexico and Brazil, requires comparisons of national product figures from before and after the export boom, and might even reveal where productivity faltered. A preguano GNP estimate is feasible, and the price index would allow for direct comparisons, particularly when improved data surface for the 1870s.⁶¹

A price series also allows us to more rigorously test explanations of Peru’s “lost opportunity” with guano. Hunt’s well-conceived model of the “rentier” economy of guano is now the paradigm. In this view, cost-price pressures were the culprit, not the enclavish nature of export-generated incomes. The demand effects released by guano rapidly reverberated in the domestic economy, and, as they did, helped release inflation. Guano-age inflation (and other factors) exacerbated chronically overvalued and inflexible exchange rates. Ultimately, this syndrome led to an import bias, a narrowing of the country’s economic base, and the crippling of potential native entrepreneurs, such as artisans and factory pioneers.⁶² By the 1860s, Peru had become an importing rentier nation, utterly dependent on the fortunes of guano. Peru had caught a “Dutch disease,” but sought no cure.

With price data and indices, we can begin to test aspects of this hypothesis. Inflation is one building block of the model, yet both its origins and its impact remain a black box. For example, was the guano era’s initial price change (1845-55), which now looks modest, truly strong enough to shift the economy to a damaging import bias? Or was this bias a more direct result of an ad valorem free-trade policy (1852. . .) in an era of plunging import prices? Combined with other price information, such a study could even advance at the microlevel, examining how cost-price pressures affected workshops or factories struggling against import competition.⁶³

If inflation emerged as an obstacle to growth, its chief mysteries lie after 1855. Once again, only a studied comparison of tradables/nontradables price indices can resolve the degree of distortion in real exchange rates, and their biases for different lines of production. Such studies have, for example, proved quite effective in understanding Peru’s path to dependent development in the early twentieth century.⁶⁴

In any case, the cursory evidence suggests that, contrary to Hunt’s model, Peruvian inflation was not severely out of line with world trends, not even in domestic foodstuffs. At the same time, the secular shift to tradables is abundantly clear, even from the price data. But whatever the precise behavior of relative prices, the roots of this guano-age inflation still must be specified. Were domestic supply constraints so critical to overvaluation, as Hunt suggests, or do domestic food prices (and the evidence of abundant supplies) suggest other pushes?⁶⁵ Perhaps profligate Peru’s recourse to overseas borrowing—which must have buoyed exchange rates above current export earnings—was part and parcel of the rentier syndrome.

On the whole, then, this price series sheds light on Peru’s path to integration with the nineteenth-century world economy and on dilemmas shared by many of the opening Latin American economies of the era. One remarkable finding is that, over these seven decades, Peruvian price movements paralleled trends seen throughout the North Atlantic economy: the high-priced instability until the 1820s; the deflation to the 1850s; and the inflation from the mid-1850s to the mid-1870s.⁶⁶

Does this rough correspondence mean we should abandon purely Peruvian causality? It at least implies that Latin American economies were well integrated with the world economy from an early date. An interesting test is whether the correlation proves weaker with colonial-era prices, and stronger as the century progressed. What will be learned about this process from new nineteenth-century indices from other parts of Latin America, including the rural zones? One also might investigate the price transmission mechanism from the world economy. It could have worked through high prices and demand for primary exports (as often assumed), combined with bullion movements, exchange rates, an expansive tradables sector, or even consumption effects and modern financial flows.⁶⁷ And, finally, to what extent did this price dynamic spread from commercial, coastal, cosmopolitan Lima to the backwaters of the Andes?

2) Social Analysis

A reliable price series can guide many forays into social history—from specifying distributive trends and class relations to identifying the causes and patterns of social protest. Prices may even illuminate the economic mentalities of an age.

On the broadest level, historians may wonder about the contrasting social effects of deflation and inflation in preindustrial settings, a theme of European historians from the Black Death to the price revolution of the sixteenth century.⁶⁸ A general proposition is that deflation tends to favor those groups whose consumption lies largely outside markets (the poor); conversely, inflation is often seen as redistributing income (through wages and profits) to wealthier classes. If wages (or payments in kind) are set by custom, the impact is stronger.

For Peru between independence and the guano age, deflation may have had a democratizing effect. It could have eased social distances by raising the real incomes of day laborers (jornaleros) and subsistence farmers vis-à-vis those of faltering elites linked to the urban commercial economy. Moreover, classic price and profit incentives to enclose popular resources (such as communal lands) were on the wane, as usual at a time of depressed prices and fragmented markets.⁶⁹ In the towns, guild monopolies revived, posing so minor a threat in a low-price era. There were many loosening social hierarchies in the postindependence era: between landlords and peasants, artisans and apprentices, masters and slaves.

The guano age, with its inflation, would have reversed trends on many fronts of class conflict. Did inflation tangibly aid the reconstitution of elites? Many factors can redistribute income during rapid growth. But it is precisely in the 1860s that one detects renewed efforts to expand Peruvian estates, some specifically for domestic food production. Apart from buoyant prices, old entails became less a burden to landlords, and new forms of credit became available at strikingly low (and plainly redistributive) real interest rates.⁷⁰ In Lima, inflating urban real estate must have become a hot investment for the rich. For urban areas, analysis might also focus on the relative class impact of imported and domestic goods. Did inflation distribute income and clout away from subsistence-level artisans, shopkeepers, and laborers, and toward elites whose marginal propensity to consume included a high penchant for cheapening imports? The price protests invariably erupted from below.

The content of popular protests (as in 1855, 1858, 1865, and 1872) always had something to do with prices. Indeed, Peloso argues that after midcentury subsistence became the most explosive people’s issue in urban politics, integral to the political formation of a modern proletariat.⁷¹ Peru’s government placated the masses with growing recourse to free trade in foodstuffs, and precocious populists (such as Mayor Pardo) seized on the anti-inflationary issue. This price series underscores these developments, by unveiling a correspondence between protests and true inflationary moments. Still, there are limits to such analysis, for Limeño crowds indeed thought in subjective terms of “just” and customary prices, i.e., of a subsistence security, without sharp fluctuations. (Historians probing the relation between inflation and protest should perhaps concentrate on foodstuffs, and on variations around the trend, where popular perceptions themselves often focus.)⁷²

Yet, before the 1850s, the main popular standard was actually the opposite one: incessant artisan demands for higher price levels through protectionism and guild monopoly. Rarely did popular groups embrace subsistence causes, not even against the Lima bread monopolies that took such flack after independence. This all makes sense amid a deflation. Between 1848 and 1858, however, the transformation of urban politics occurs. Craftsmen lost out to Manchesterian free trade, watched their guild powers vilified and undermined, and joined the noisy chorus for inexpensive wage goods.⁷³ This was not just the result of effective liberal propaganda, which by then was borrowing newfangled and populist Manchesterian ideas. The inflationary shock of the 1850s palpably helps to explain the shift to free-trade politics and policy, as well as this changing content of popular mobilization.

These matters are all part and parcel of the distributional question of the guano age. With a price index, this social question (as it was called) can become far more precise. Real wage rates, for example, could be explored thoroughly in Lima. Spotty data reveal that urban jornales remained stationary at the traditional 6 reales from independence through the 1850s; the daily wage then rose to 7 and 8 reales over each of the next two decades. Therefore, in constant 1830 reales, wage growth was on the order of 17 percent by 1840-42, or 19 percent by 1850-52—due simply to deflation. Wages then fell 7 percent below 1830 levels by 1860-62, and 19 percent below them by 1871-72.⁷⁴

The striking finding here is that, given the largely stationary urban jornal, real wages dropped nearly a quarter (22 percent) during the first decade of the guano age. And this measure, of course, does not even consider the scourge of unemployment among Lima’s teeming population, estimated at roughly 16 percent by 1858. Small wonder Lima’s destitute violently took to the streets in December 1858 against this “fictitious prosperity.”⁷⁵ Among the popular demands were lower rents and lower food duties, but not higher wages.

Does social custom help explain this adverse wage-distributional effect during inflation? Commercially oriented elites cope better with inflation, as they are accustomed to bargaining over prices as a way of life. For whatever reason (including adscriptive prejudice and barriers), popular groups seem to lag in adopting a haggling attitude on wages. In Lima, for example, the first recorded strike over wages occurred only in 1869—and was organized by construction workers, aptly enough, while tearing down the city’s overflowing walls. But this is a more general query. In parts of rural Mexico, for instance, a stationary peon wage has been reported for most of the nineteenth century.⁷⁶ Surely that was a boon to hacendados, and a plague on workers, as price levels inched upward during the century.

Rural Peru had its peculiarities. Scattered evidence suggests that wages on coastal plantations rose dramatically in the late guano age: from a 3.5-real norm in the mid-1850s to anywhere between 7 and 20 reales by the early 1870s.⁷⁷ This was due to Peru’s chronic agricultural labor shortage—the so-called falta de brazos—which meant, in reality, real 20 to 200 percent pay raises for workers, not counting the premium often afforded by food rations in kind. These estate costs no doubt added to high urban prices. Only landowners with the foresight and means to acquire coolies (with fixed long-term wage contracts) blunted the problem.

But why were real urban and rural wages moving in opposite directions after 1850? Price history suggests that migration patterns are worth exploring further in nineteenth-century Peru. Contrary to our image of Peru as an immobilized land, some 37,000 internal migrants flocked to Lima during the 1850s, becoming more than a third of the population. This human wave surely contributed to the dampening of urban wages during the guano age. Yet, what brought migrants to a reviving Lima, and not, according to all witnesses, to the expanding coastal plantations that were offering premium wages by 1860? One wonders about discrepancies between low provincial and inflated capital price levels—was this an initial (if chimerical) pull factor for those mainly seeking remittances back home?⁷⁸ It may have been easier to see high prices (or nominal rates of pay) than to calculate real wages. On the other hand, the Andean migrants’ aversion to work on high-wage coastal estates seems even more perplexing. Perhaps theirs was not a price calculation at all, but a cultural aversion to oppressive estate regimes and to laboring along with oppressed minorities like coolies and blacks. These are among the more complex social history questions raised by the simple history of prices.

3) Political Analysis

State building has recently emerged as a pressing theme in the history of Peru. However, we still have little grasp of the real growth of the guano-age state, though this was in effect the formative stage for Peru’s modern national/administrative state.

In current expenditures, state spending swelled from 4.5 to 48.8 million pesos between 1847 and 1872, or over tenfold! And this estimate excludes all external and internal debt transfers. Calculated in real terms— the resources the state actually commanded—that expansion was more in the range of 5.25 times, still an impressive record (or waste). The guano-age state’s real revenues increased 4.7 times. Moreover, real growth was steadily upward; roughly 55 percent occurred before the elevated price level of the 1860s, and 45 percent thereafter.⁷⁹ So, inflation per se did not halt the expansion of the state.

Disaggregated trends should also be revised. One example is the statistics produced by the indefatigable Hunt on empleomanía, concerning the traditional allegation that the state threw away its income in low-productivity political employment. In the course of the guano age, if measured in constant prices, average salaries of senior bureaucrats such as ministers and prefects actually decreased some 44 to 59 percent. Global spending on civil servants and military men (including their pensions) went from 3.6 to 7.3 million pesos between 1847 and 1870, an apparent rise of 103 percent. In real terms, this meant a tiny change of 5.9 percent.⁸⁰ For these elites, the guano age was a surprisingly fictitious prosperity, too. To be sure, such analyses overlook other pertinent factors in income concentration, corruption, or influence, e.g., the existence of multiple prebends for elite family networks. But the clear point, as Hunt already suggested, is that bloated employment was not the core difficulty with the expanding guano-age state.

Social theorists have often postulated a close relationship between state building, extractive capacities, and inflationary cycles.⁸¹ The guano-age state, however, was largely “autonomous” in that it owned and borrowed on an independent source of income, its guano monopoly. By 1855, the enriched treasury abandoned any dependence on direct taxation, and the state lacked the will and means to return to tribute and other taxes when deficits widened and costs rose. By the 1860s, greater than 70 percent of revenues came from sales of bird dung.

Major fiscal struggles thus did not revolve, as so often happens, around peasants, regional groups, or other taxable targets. Apart from its sheer wealth, this fiscal freedom is what probably allowed the state the ability to expand so rapidly, while avoiding the social conflict endemic to such processes during inflations. A revival of Indian tribute, for example, would have sparked sharp resistance, for quotas would have had to exceed customary levels.⁸²

What, then, was the state’s response to its perceived crises in real income? Peru turned outward to massive foreign credit in the 1860s and 1870s, amassing Latin America’s largest debt on the London market. Historic proclivities, a mania for huge projects, a strong credit rating, and wavering guano sales all contributed to this outcome. Yet, one wonders also whether the inflation of the era became an incentive to this growing, and ultimately disastrous, recourse to overseas loans.⁸³ A relationship between inflation and loans is discernible in a typical fiscal ratchet-effect during Peru’s episodes of war. In tandem, debt, spending, and inflation leapt upward, and the size of the public sector never returned to its previous levels.

And did overseas loans, in turn, just contribute to vicious inflationary borrowing cycles? This state appears to have done little, given its extractive strategy, to dampen inflationary demand—whereas those which rely on taxation or local borrowing do, inadvertantly and now purposely, as they expand. The guano-age leviathan appears inflationary in multiple senses of the word.

4) Regional Analysis

Some of the food prices in this study provide clues to regional agricultural development, for example, the long-term shift in the price of lard coming from the valley of Chancay. Were there any productivity gains at work? And did true supply inelasticities exist with some of the products of more volatile price, such as rice or beef? We still have little grasp of the growing impact of food imports on local producers, or of estate decisions to invest for urban or overseas markets.⁸⁴ Continuous price data help here, but we surely need prices from estate and regional records as well.

Regional economic integration, often studied under the rubrics of “internal market development” or “commercial penetration,” is a burning issue in Peru’s modern historiography. The old image of an utterly disarticulated, nonmarket, immobilized, or “feudal” nineteenth-century Peru is fast fading. On the other hand, the extent of Peru’s market integration is not directly addressed. Some historians still mistrust quantitative work, precisely on the ground that for fragmented regional economies any uniform statistical series will prove misleading.⁸⁵

True, a spurious precision can infect economic history. But the core question here is actually empirical. To the extent that historians collect regional price data to complement these from coastal Lima, we could test the varying propositions that national market integration was nonexistent, decreasing, or increasing during the century. Cursory data suggest that Lima price levels were “higher” than those of provincial centers, though it is the time-series correlations between regional prices that matter in this problem.⁸⁶ With such studies, historians could resolve the possibilities, and limits, to further quantitative and market-type analyses in this preindustrial setting.

As discussed, Peru’s coastal economy appears to have been deeply mired in world economic currents, and we need to know how these washed into the sierra. The most obvious need is to gather regional prices, and explore the range and depth of markets. However, regional concerns also raise the cultural/spatial dimension in all price studies. Lima’s consumption patterns, which most resembled those of the West, formed the basis for this study’s particular cost-of-living index. Regardless of findings on market influence, historians may justifiably decide that, culturally and ecologically, the various peoples and corners of Peru were worlds apart. A coastal index cannot suffice for all problems. Regionally based indices, derived from local consumption patterns, would be particularly apt for assessing the impact of price movements on welfare.⁸⁷

Let me mention, in closing, the spirit of such research suggestions. Several decades ago, a venerable social scientist characterized the process of economic development itself as proceeding from a ferment of linkages, bottlenecks, and trickle-down effects; development is not a “balanced” process, nor from anyone’s master plan. An awkward price, for example, might act as a stimulus to growth across many sectors.⁸⁸ Much the same can be said of progress in historical knowledge, which rarely flows smoothly or from agendas. Thus, more important than any of this study’s rough findings are the spurs they may provide to our rapidly developing field.

Price Data, 1800-73. In Current Reales (8 rls = 2 peso; 10 rls = 1 sol)

Author notes