(Bloomberg)—As U.S. consumers click their way to checkout, it’s reshaping a big chunk of the nation’s workforce.  A new Federal Reserve Bank of New York analysis digs into who wins and who loses when shoppers migrate to the web, and that’s the first study in this week’s economic research roundup. We also include a look into how gentrification and crime relate, summarize a skeptical view on the Fed’s understanding of inflation, and outline the fringe benefits of education. Check this column each week for the latest in economic research from around the world.  Online shopping and retail jobs More than 10 percent of retail sales are now made online, up from 2.5 percent in the mid-1990s, and the job market is feeling the fallout. Since 2013, department stores have cut about 80,000 net jobs as nonstore retailers, the group that includes online, added about 100,000, according to New York Fed researchers. The trends “show no sign of abating” and may actually be intensifying, authors Jason Bram and Nicole Gorton write.  Given that nonstore retailers are adding a lot of jobs, this may seem like a wash. Still, there are clear winners and losers. For one thing, the places bleeding stores jobs aren’t the ones gaining online work: between 2012 and 2016, about 75 percent of the counties the authors studied lost department store jobs, and four-fifths of those counties lost out on net. First among the job gainers? King County, Washington, home to Amazon.com. The required skill mix is also different. The average online retail job paid slightly over $50,000, while the average department store job paid just $20,500 – so if pay is a proxy for skill level, we’re shifting employment up the ladder.  How Is Online Shopping Affecting Retail Employment? Published Oct. 5, 2017 Available on the Liberty Street Economics blog Finding balance with services trade Future liberalization of services trade could both generate increased gains from trade and help to reduce global imbalances, Bank of England researchers argue in a blog post. Goods trade has liberalized much faster than services trade since the 1990s, so “those countries that are focused on exporting goods have seen big increases in exports, but those focused on services haven’t.” That could change if services trade increases, boosting countries like the U.K. and the U.S. that are more service-heavy economies. In fact, liberalization in countries with the highest barriers – including India and China – could narrow the U.K.’ s current account deficit by 0.7 percentage point, they estimate.  Mind the gap: Services trade liberalisation and global imbalances Published Oct. 10, 2017 Available on Bank Underground website Links between gentrification and crime reduction In Cambridge, Massachusetts, a sudden end to rent control in 1995 caused overall crime to fall by 16 percent—about 1,200 reported crimes annually— a drop driven by property crime, according to new research by Massachusetts Institute of Technology economist David Autor and co-authors. That saved would-be victims about $10 million in 2008 dollars, they estimate, and capitalizing that suggests it probably caused about 15 percent of the increase in local property values. The takeaway? Crime avoidance actually causes further price appreciation, and “improvements in public safety are therefore an economically significant part of the gentrification process.” Gentrification and the Amenity Value of Crime Reductions: Evidence from Rent Deregulation Published October 2017 Available on the NBER website Daniel Tarullo questions everything Many Fed-watchers have questioned whether the central bank is hanging onto an outdated inflation model simply because they need to: if the tried and true relationships don’t stand up, they’re setting policy without much theory to back up their decisions. Add former Fed Governor Dan Tarullo to the list of skeptics. For background, the Phillips Curve – a long-standing relationship where lower unemployment pushes up inflation – doesn’t seem to be working like it used to. Inflation has remained tepid even as unemployment has plummeted. “The relative unhelpfulness of these past correlations leaves the FOMC without a working theory of inflation,” Tarullo writes in a Brookings Institution paper. “This, I think, is highly disconcerting for a profession that it is very attached to modeling.” Tarullo goes so far as to suggest that economists may be dismissing data that don’t fit with their narrative as transitory so that they can stick with their view. He also argues that the Fed simply doesn’t know if well-anchored inflation expectations will actually pull prices back toward its 2 percent inflation goal, which it’s been undershooting for months. His conclusion: the Fed should be looking at models and traditional relationships, but should rely on actual data when setting policies. As he sums it up, “the problem lies more in the coefficients than the concept.” The paper is one opinion, but it concisely sums up some of the biggest debates in U.S. monetary policy right now.  Monetary Policy Without a Working Theory of Inflation Published October 2017 Available on the Brookings Institution website