You don't get anything tangible from that noisy data. IT has been taking the brunt, and you only see labor market changes after they happen by a significant lag. Its premature to do this type of study and try to discount what people on the street who are being impacted are saying.