July 2017 Connecticut Sentiment Index

The July Connecticut Sentiment Index barely budged from its June level, inching from 57.7 to 57.9.  By contrast the University of Michigan Survey of Consumer Sentiment Index declined (-2.1%) between June and July. Survey of Consumers chief economist Richard Curtin attributes the decline primarily to fading expectations of a continued Trump bump.  Recall that the Michigan Survey is a national telephone survey whereas the CSI is built from mining Connecticut web sites.  Still, as for the CSI Connecticut, its impossible to pinpoint any specific reason for our flat result. There is quite a bit happening in the state lately that would send the positive and sentiment word count in either direction. 

For instance, among all the news coverage this week there was an announcement - amidst much gleefulness - around a recent CNBC Survey results which ranked Connecticut 33rd in the nation (the gleeful part come from the fact that we improved 10 points in the ranking when compared to the results last year).  [see it here] Yet, if you do look through the hyperlink, you will notice that the CT News Junkie has a July 12 piece on Standard & Poor's downgrading of Hartford bonds to junk status sitting exactly alongside the CNBC survey. [article here; front page of Junkie here]  Ratings are assessments of a city's credit-worthiness and serve to gauge the city's likelihoods of bankruptcy. 

These contradictory announcements are quite perplexing for establishing net sentiment purposes; and probably great pickin's for those seeking confirmatory evidence for their priors.

So we took a look at the individual positive and negative indexes that constitute the Connecticut Sentiment Index to try to

draw something out.  That result is displayed in the 2nd graph.  Positive counts declined, negative counts declined between June and July.

I'm stymied; write to me with your thoughts. 


About the CT Sentiment Index

The Connecticut Sentiment Index (CSI) is based on a count of positive and negative words appearing in the online pages of 12 regional and local newspapers. We scrape in the evening of the 15th of every month - or thereabouts.  We try to publish on the day the University of Michigan Survey of Consumer Sentiment.  The CSI sentiment classification is obtained using the Hu & Liu Lexicon of Opinion Words in English. The index is the ratio of positive to the sum of positive and negative words, multiplied by 100.

Email me when people comment –

You need to be a member of UNH Economics Collective to add comments!

Join UNH Economics Collective