Unravelling the influence of environmental and socio-economic factors on historical grain price variations in Medieval and EarlyModern Italy
Summary
Grain Price (GP) volatility is central to pre-industrial economies – in the past, governments have fallen because of their inability to manage food crises, many of which arose from grain shortages. In this thesis, I present an in-depth analysis of factors that have caused GP variation across the Northern-Italian city-states of Milan, Venice and Tuscany, from the 14th to the early 19th centuries. Following Adam Smith’s logic of supply and demand controlling prices, changes to weather should play a decisive role on GP variation by affecting the harvest success/failure. Starting from this assumption, I first explore temperature as the key determinant of grain price volatility across Northern Italy, both looking at wider European temperature through the Luterbatcher temperature reconstruction, and a more local Alpine tree-ring record. After using de-trending methods to remove the effects of long-term inflation on the economy, I conclude that direct correlations between GP and temperature, across Italy, are low (r<0.1). Never-the-less, a similar analysis on Stockholm, Sweden, shows a stronger relationship (r=0.2±0.1, de-trending method dependent). Granger-causality only produced positive results for Sweden, and I argue that failure in correlation in Italy is because of Europe-wide trade, and Italy’s uniquely advanced provisional mechanisms. In addition, by applying Empirical Mode Decomposition I find periods of sustained high (≈5 years) GP to correspond with periods of intense war, social unrest and plague, also displaying a cyclic behaviour. I show that economic cycles either do no play a role in GP, compared to social unrest and war cycles, or that economic cycles have a subtly different mechanism than today. Ultimately, I conclude that GP variation is influenced, first and foremost, by population and global silver supply, then subsequently by: market integration, plague (feeding back into population), war, social unrest, and then weather. However, I further conclude that a linear regression model based on population and temperature can explain significantly GP volatility during times of peace (r=0.63), as reflected in the situation of 18th century Milan.