Hi Anna,
Thanks for the helpful “Presimetrics” book tip when contributing to my “Who Do You Trust, Baby?” post. No, I am not familiar with it. I was not aware that the book even exists.
Sorry for the delay in my reply; I was on holiday for a couple of days.
You are absolutely correct in reading my likes, preferences, and tendencies. I would not want to play poker against you.
I am a sucker for anything statistical. When I was a kid, I taught myself arithmetic long-division at home so that I could calculate baseball player batting averages. I was of a sufficiently skeptical bent even at a young age to challenge daily newspapers. I needed to verify my heroes’ batting averages independent of the reporting.
I have accumulated enough experience dealing with the statistical discipline to understand its possible benefits and shortcomings. The quality of the data depends on the planning and care exercised in collecting it, and on a fair and unbiased interpretation of what it potentially tells us.
Sometimes interpreting data presents a few subtle challenges. Action time lags (in engineering hysteresis phenomenon), disparities between short-term and long-term effects, complex feedback loops, and even identification of who or what were the decisive protagonists or factors are often difficult to assign properly. The economic community still debates the causes and cures of our Great Depression over 80 years ago. The blame game seems to be an endless dispute that persists forever.
Although I referenced a few statistics that correlate market returns and political party control, I am not sanguine with the relationships implied. These data sets are longitudinal by necessity, and the people that populate these parties are in constant flux. Political party policies, platforms, and affiliations are not immutable over time. To illustrate, voting Blacks were solid Republican after President Lincoln in the late-1860s, and switched to solid Democrat after President Lyndon Johnson in the mid-1960s. Shifting allegiances are everywhere.
The Democratic party that swept Andrew Jackson into the White House in 1828 is not today’s Democratic party. The Republican party that swept Abraham Lincoln to the Presidency in 1860 is not today’s Republican party. Additionally, although political parties gravitate towards a consistent government philosophy, Presidents are more free to pursue their own specific world views. “It’s good to be king.” Therefore, although statistics do usually light the way, in this instance it is a dim and flickering flashlight.
And please recall the most serious caution about any statistical data set and analysis: Statistics are simply a backward looking record of history. It is a nice statistical tool to organize and present complex measurements. It shows what happened in the past as a guide to possible future events. But it does NOT predict the future, especially in the social realm.
Scottish philosopher David Hume observed that there is no universal law of nature that the future must mimic the past. In all instances future prospects may resemble past circumstances, may be similar in many aspects, but are never absolutely identical. The failure at exact replication almost always generates unexpected deviations and a failure to forecast accurately. Just witness the dismal reputation that economists have earned for their sorry record of poor projecting. Societal things in the real world are seldom linear.
The “Presimetrics” book that you referenced looks like a reliable source that was written by a seasoned and serious economist. I’ve only examined the Table of Contents. The chapter headings imply subject matter that is in my wheelhouse.
Chapter 1 deals with real GDP growth rate per capita. I believe that’s the single most significant economic measure that defines our prosperity level and gains, and probably, in a very general sense, our happiness. The book should be stimulating for a statistical nut like me. I expect it to illuminate and it is likely to be somewhat provocative. Good.
I’ve added “Presimetrics” to my “must” read list. I expect I will not get a copy until I return from a 3-week vacation.
Thank you once again for the alert.
Best Wishes.
Comments
Of course, studying correlations for short periods of time is what it is - riddled with problems. It's not a good data base. The last two Presidents are good examples. They share a similar first year economically - a falling stock market that originated before their watch. Presidents don't establish their first year budgets but they do establish the next guy's first year budget. Just two examples where time lags exist. I think Mike did a good job with the data. However, for the most part he stuck to simple correlations so they are what they are. His book, for people like you, is a lot of fun and a few laughs especially when the stats seem to indicate other than what you think or believe.
I am from a science and statistics background (modeling really). Economists just mystify me and I enjoy things that make me smile.