‘Esta breve historia’ – this brief history – these are the opening words of the introduction to Bruno Seminario's extraordinary book presenting his original estimation and analysis of the Peruvian national income accounts back to 1700. Surely there has rarely been so inappropriate a beginning, since it needed 1,298 pages for the author to recount and present all the scholarship set before us after many years’ work. The book reconstructs the national accounts for Peru back to 1700, and makes estimates on an even longer period. To do the reconstruction, regional and sectoral estimates are constructed and presented. Not content with this, the author presents an analysis of enormous breadth, covering European and sometimes world history, to derive a comparative picture over the long run. The methodology is thorough, and sophisticated in its use of filtering and smoothing techniques. The author's knowledge of several huge literatures is extraordinarily impressive. The effort to determine the evolution of the terms of trade, the real exchange rate and other relative prices, is valuable and impressive, and covers territory that the ‘Bible’ (Angus Maddison) did not attempt for Latin America (indeed, he could have been more critical of Maddison). His concern to evaluate the impact of natural disasters and epidemics results in new and original work and is to be welcomed. He frequently checks his findings against common sense and others’ findings.

Inevitably, in a work of such vast scope and detail, there are many points to be debated. There will be huge debate not only over the methodology but also over the interpretation. My own largest doubt concerns domestic institutions and the political economy of the interaction of the growth path and domestic politics, policy and institutional shaping. The author finds that cycles are crucial to Peru's performance and develops a worrying counterfactual that without wars and natural disasters Peru would have grown faster than the United Kingdom (p. 1231) or would have equalled Europe (p. 223). In other words, external events appear to be entirely responsible for Peru falling behind. I would suggest an alternative interpretation of his data. It may be that either or more probably both the vulnerability to crisis and the slow speed of recovery are at least in part caused by the impact of the preceding boom period on relevant institutions. That negative impact may take the form of the distortion of institutions and/or policies, or simply the absence of development. Thus the guano boom seriously affected tax collection, diverting finance to government and bureaucracy centred on Lima, and produced exchange rate and tariff impacts that damaged incipient industrialisation. It took external forces including the collapse of silver to generate recovery in the 1890s. When we come to the modern day, the author rightly mentions increased stability by the 1990s and better management mechanisms as elements helping to moderate the impact of the cycle, but surely there were other forces at work: my own analysis would be that the impact of the debt crisis of the 1980s had to do with the extent of the structural supply side crisis in exports and the weakness of institutions to generate internal dynamism. Domestic institutional dimensions in their turn are not a sufficient explanation: in the 1930s, for instance, contingent aspects outweighed structural institutional weakness – hence the relatively rapid recovery (the early revival of the international market for cotton, opportunities for easy import substitution and for the expansion of small mining).

Turning to the organisation of the book and the methodology, I found the order of presentation challenging. The reader is given over 400 pages of vigorous international comparisons and discussion before we have a chance to appraise the quality of the author's estimate of Peru's growth which he is placing boldly in a worldwide context. I found I needed to form my own view first on their solidity and the possible biases.

Once we arrive at the explanation of the methodology, the approach chosen is sound. The author derives his estimates from the output side with a careful base in population figures. As always, it is agriculture for the domestic market, and in particular subsistence production, that is most problematic, and this is where some of the least plausible results appear, as he acknowledges, for example, the sharp changes that result in the structure of PBI in the calculations on page 1039. To have the share of agriculture varying in different directions by as much as 50 points over a relatively small number of years is counter-intuitive and reinforces concerns about the methodology of the estimates. Again, urban/rural per capita is given in Figure 1v –14 (p. 498) and shows as he says a surprising dynamism in agricultural income per capita through the twentieth century, while urban income per capita nearly stagnates. However, we need to evaluate what part of this result was built in by the methodology, and this is not made clear – indeed there are many reasons to be concerned about the calculation of agricultural income. Agriculture for the domestic market is the difficult one to estimate, and the author (rightly in my view) dismisses Richard Webb's national income calculation where production of food is deduced from income assuming distribution of income constant. But the author still relies on the underlying assumption of this methodological approach (a population increase generates food production) for the early twentieth century and for 1949–70 he uses ‘La Renta Nacional’ – figures Webb discarded with good reason. It would also have been insightful to have examined the role of food imports – running at some 16–20 per cent of all imports from the 1920s on, concentrated in urban food consumption and systematically favoured by government pricing and policy.

In general, the clarity with which the workings and sources for his estimates are presented is one of the most notable and praiseworthy aspects of the work. The detail of the original estimates is painstakingly set out, enabling us to form our own judgement as to the plausibility of the results (with only one or two exceptions). The author has used every possible piece of research by scholars over the last 60 years to piece together his estimates, which will of course be controversial, but we can explore them and form our own conclusions. In a work of this size, of course, it always seems ungrateful to ask for more, but I would have welcomed more informal sensitivity analysis; it would be invaluable to the reader to be guided as to which assumptions really make a difference to the trends over time. On page 834 there is a welcome presentation of the extent to which different sectors’ estimates depend on the solidity of the indicador coincidente but more of that sort of evaluation would be valuable. A good example is the selection of Cook's estimate of population. This is commonly accepted as the best available, but it would help the reader to know why it is thought to be solid and how much difference it makes to vary the assumptions behind it within plausible ranges.

The remarkable cover to this remarkable book portrays Hieronymus Bosch's vision of the Way to Heaven, painted in the early sixteenth century. The good Christians are being transported by angels into the tunnel of light which in his vision leads to Paradise. A curious choice of cover, though eye-catching – perhaps more apt is the scribbled Latin tag from Bosch on the back of one of his paintings: ‘Pitiful is he who always makes use of the inventions of others and never creates anything himself.’ This is a book of unusual creativity and scope, and will provide provocation and food for thought for scholars for many years to come.