A spatial analysis of suicide rates in Tennessee

Purpose - Suicide rates at the county level may be influenced by various macro-level factors emanating from surrounding communities. The purpose of this paper is to measure the extent of geographic spillover effects from socioeconomic factors on suicide rates. Design/methodology/approach - Using county level, panel data over the years 2002-2007, a spatial dependence function is applied to measure how socioeconomic factors within surrounding counties may influence suicide rates in neighboring, adjacent counties in the state of TN. Findings - A negative correlation is identified between divorce and suicide rates within each county, which seems to contradict traditional beliefs that these variables move together. Also contrary to expectations, unemployment and suicide are not significantly related within counties. When the analysis measures the spillover effects of surrounding communities, county-level variables such as income, crime, and church density are significant. And although the signs on income and crime are consistent with past findings, church density counters research expectations. Practical implications - These results suggest that certain socioeconomic factors influence suicide in surrounding communities, while others do not. Incorporating spatial analysis into future research on suicide and mental health may assist practitioners with suicide prevention. Appropriate prevention policies should be designed and implemented at the local or regional level. Regional differences make broad policies based on national data inappropriate for local areas that differ from national norms. Originality/value - Studies on the external links to suicide and suicide rates have made significant contributions to raise the understanding about mental health issues. Very few, however, have directly employed research methods to capture spillover effects when the study encompasses spatial elements.