Keshmeer Makun

In 1989, in macroeconomics lectures in my Melbourne university MBA, I couldn't believe how devoid it was of any connection to reality. I cannot believe that we are now in 2025, and a handful of rational economists are still struggling to be heard! rwer.wordpress.com/2025/05/08/the…

25% tariffs on Mexico & Canada + the 10% tariffs that are already in effect on China, could cost the typical American household more than $1,200 per year—& this analysis doesn't include the additional 10% on China recently announced. #PIIECharts More: piie.com/research/piie-…

El Nino's extreme weather is disrupting food production worldwide, forcing countries to rethink trade. Droughts, floods, and heatwaves are impacting crops from corn to cocoa. #ElNino #FoodSupply #GlobalTrade

"While there is extensive support for the budget based on certain policy measures ... achieving a sustainable fiscal position and increasing economic growth remains a challenge for Fiji," say Keshmeer Makun and Janesh Sami from @UniSouthPacific. devpolicy.org/fijis-coalitio…

EU prepares a two step strategy to deal with tariffs if Trump wins

Trade between developing economies has significantly increased, transforming trade patterns. It grew 9.7% annually, surging from less than 10% of global trade in 1995 to nearly 25% by 2022, reaching US$ 6.1 trillion. More details: bit.ly/49Q3lIW #WTOat30

Providing homeless people with housing: - reduces crime - increases employment - improves health - does not increase reliance on social benefits. 80% of costs are offset by the benefits in the first 18 months:

Robert Solow won a Nobel Prize in 1987 for his analysis of how technology drives economic growth in developed nations. He died Thursday at the age of 99 time.com/6550070/robert…

This @IMFNews paper has a great overview of different public debt databases and their coverage: imf.org/en/Publication…

Mean Absolute Error is the simplest evaluation metric. It tells us how far the predictions are from the actual values, on average. Let's break it down to Grandma's level with an example. 🔽 Consider a linear model with 5 datapoints. (n=5) We need 2 sets of values to calculate MAE: • Set 1 contains the actual values. [3, 4, 5, 6, 7] • Set 2 contains the predicted values. [2, 5, 4, 6, 8] The error shows how far the predictions are from reality. In math 'how far' translates to difference. So we want to know the difference between actual and predicted values. In MAE we consider absolute values. Why? Consider the first 2 pairs of points. Actual values = [3, 4] and Predicted values= [2,5] For the first point, the error is 3-2=1, while for the second, the error is 4-5=-1. If we sum these errors we get zero. This would tell us we have no errors, and the model is perfect. But we see it is not the case. We need absolute values to avoid cases where positive and negative numbers offset each other. The absolute errors for our example: [|3-2|, |4-5|, |5-4|, |6-6|, |7-8|] = [1, 1, 1, 0, 1] As the formula says, we then sum these errors. Sum of absolute errors: 1 + 1 + 1 + 0 + 1 = 4 And finally, since we are talking about mean error, we need to divide by the number of observations. Mean Absolute Error: 4 / 5 = 0.8 The result means that on average, the predictions are 0.8 units away from the actual values. If we want to find a better model, then we need one that has a lower MAE than 0.8. ___ That's it for today. I hope you've found this Tweet helpful. Like/Retweet for support and follow @levikul09 for more Data Science content. Thanks 😉

China now has two universities in the top 15 in the world, and 13 in the top 200 according to @timeshighered. We need to be engaging the top institutions of the future, not decoupling from them. timeshighereducation.com/world-universi…

First year grad student cleaning data be like twitter.com/FunnyXPost/sta…

Why are interest rate rises not taming inflation? on.ft.com/3rdahiC

I can attest to this, "China went from 80% extreme poverty in 1980 to NO extreme poverty 40 yrs later" Jeffrey Sachs I was in China in 1986 and food was available but scarce. Although there was plenty of watermelon and white cabbage piled on carts, stored in alleyways and communal courtyards. A friend who was in Dalian in 1989 saw these white cabbages being buried in long troughs dug into the ground as a way to preserve them for the winter ahead. People back then remembered the 1960s when state-rationed turnips, leeks and cabbage sustained millions during Mao's Great Leap Forward policies. We should take joy in China’s accomplishments. Not fear from it.

China🇨🇳 will remain an economic powerhouse, expected to contribute one-third of global growth this year. As some European business people have told me, investing in China is an investment in the future.

The FAO World Food price index declined sharply in March. Food inflation is set to come down in the next months.

Hi #EconTwitter! Looking for a graduate #Econometrics course, covering lots of stuff, including model/moment selection in high dimensions? Check out these very enjoyable slides and notes by Francis DiTraglia @economictricks (@OxfordEconDept). Very cool material!

Happy Monday #EconTwitter! Are you an economist willing to learn how to code in 𝗥? You should check out this series of videos by @nickchk (@SeattleU), which spans from introductory lectures to advanced stuff. It’s definitely an all-time favorite!

NEW WORKING PAPER titled "Diminishing Gains from Trade: A Macroeconomic Analysis" is available at ssrn.com/abstract=43641… One sentence summary: The gains from trade decrease with trade openness when country-specific trade elasticity estimates are used.

China is far ahead of Europe and the rest of the world in new offshore wind installations. Really pushing the renewable energy stuff, so that Chinese industry has a massive advantage (see also solar)... via @scienceisstrat1