Anders Milhøj

A Conditional Variance Model for Daily Deviations of an Exchange Rate

A model for the distribution of daily deviations of an exchange rate is suggested. The distribution is Gaussian with a variance that depends on previous deviations. The model is applied to the exchange rate of the U.S. dollar to special drawing rights.

Causes of deforestation in southwestern Madagascar: what do we know?

Abstract Causes of deforestation are discussed in the case of southwestern Madagascar. The article has two objectives. First, the idea is to discuss the usual linear regression approach to determine causes of deforestation. Secondly, the intention is to determine the causes of deforestation in our case study. Distinction is made between direct causes (agriculture, wood collection, and pasture land) and indirect causes (migration, export prices, property rights, and government policies). In outlining the strategies of the peasants, the article ends up with an estimation of the value of agricultural land vs. an estimation of benefits derived from utilisation of non-timber forest products from primary forest or scrub (value of medical plants and roots used as staple food).

The sale of alcohol in Denmark – recent developments and dependencies on prices/taxes

How do prices affect the choice of types of alcohol in Denmark? We study the Danish sale of alcoholic beverages in a time series framework. First, we look at annual data from 1980 investigating the hypothesis of a fairly stable level of sales. We conclude stationarity of sales and we also find that the income elasticity of total sales has been zero. Second, we analyse the composition of the alcohol sale between beer, wine and spirits in a multivariate model conditional on the development in prices. For this analysis we use Johansen cointegration techniques. Again we test that income can be omitted from the model and we use the model to derive the effects on the composition of alcohol sales of three different sets of changes in the alcohol taxation.

Vulnerability in north-central Vietnam: do natural hazards matter for everybody?

This article examines changes in livelihood strategies in response to flooding. It does soon the basis of a household survey which was undertaken in three provinces in north-central Vietnam. All households in the survey were regularly affected by flooding, but only poor households experienced long-term negative effects. The research showed that a high impact of natural disasters is correlated with decreases in income over time. As the disaster relief offered by the authorities is marginal compared to economic losses, some households react by increasing off-farm incomes (including remittances from overseas-migrated household members). We observed that the poor households do not switch to off-farm income strategies in response to income losses; this is perhaps because they have inadequate skills. However, on average households in our survey are becoming richer over time, despite the impact of flooding in the provinces. The article ends by looking at the vulnerability–resilience debate concluding that the poorer households could enter a vulnerability loop, unless new strategies to cope with natural hazards are suggested.

The Market Share of Wine in Denmark

This paper analyzes the position of wine in the Danish alcohol market from 1978 to 2008. During this time period the market share of wine has doubled, assumingly due to a tax-induced change in relative prices as well as behavioral changes. This paper models the yearly observed market share of wine by unobserved component models which allow for adaptive trends including regressions with time varying coefficients. I find that the market share of wine has a steady upward trend even if it was temporarily broken by a tax reduction for liquor in 2001. (JEL Classification: C22, D11, L66)