A random sample of 42 locations were selected. For each location, the 2013 per capita spending on a specific item was determined and an index of "level of urban development" was created. Each location assigned a value (where 0 indicated a completely rural & 100 a completely urban area). A simple linear regression relating expenditure to this index was performed. The estimated intercept and slope were a = 1.96 and b = ?0.0211, the sample correlation coefficient between the two variables was r = ?0.888 and the MSE from the regression was 0.1603.
What is the predicted per capita expenditure for locations with an urban development index value of 50?