Þetta hefur nú engin áhrif á ákvarðanatöku ríkisstjórnarinnar held ég, þetta er bara stjórnarandstaðan í málþófi að röfla um ekki neitt og meirihlutinn að leyfa þeim það fram á nótt.

Most of the outliers within countries are surveys that use unusual response scales, and even then, the trends are mostly the same. Dichotomization also produces essentially the same trends as using the full scales...

Thanks for reading! Find the full paper and SI here: doi.org/10.1017/S000...

One of my main take-aways from doing this project was the remarkable consistency between different measures from various survey projects in terms of trends and between-country differences. Here's trust in government in WENA. Some differences, sure, but overall... they are all on to something!

There are even more nuggets in our Supplementary Materials (142 pages)!

There you can find the cleaner graphs shared above; descriptive trends; results from Stimson's method; MLMs; information on weighting and scales and more (nerd-)fascinating stuff!

But trust has also been high in the Nordics (except Iceland!), New Zealand and Switzerland - suggesting that we can look to democratic countries for potential solutions to the trust crisis. Maybe part of the answer is more money! But also equality, vibrant democracy, social cohesion, etc...?

Similarly, (reported) trust has been rising in Russia since Putin took power (in 2000) and has been stable or rising in Turkey. The very highest levels of reported trust are in China and Vietnam. High levels of trust in surveys don't tell us the whole story about the state of a country...

In Eastern Europe and Central Asia, trust mostly appears to have declined in the 1990s - when many of those countries were newly independent. In Hungary, trust declined dramatically until 2010, when Orbán was re-elected and started dismantling liberal democracy. And then trust started rising!

The effects of the 2008 financial crisis are also clear in our data - but mostly in Europe. Esp. in Greece, Cyprus, Spain, and my own native Iceland. Trust also feel sharply in Italy, Portugal and Ireland in 2008, but appears to have (mostly) recovered in those countries... (graph shows parliament)

There's also an interesting non-linear trends in many countries in Latin America, where trust in representative institutions was on the rise until about 2014, after which point it started declining sharply. I'd be interested to hear people's take on this, as I'm no expert on politics in the region!

Although fairly stable overall, we also saw that trust in the legal / judicial system appears to have declined a bit globally in the 1990s. And we see clear declines in this measure in the Latin America region - especially in Mexico, Venezuela, Chile, El Salvador...

One of the findings that most surprised me was that trust in the police had been rising globally by about 12,5 percentage points! But it's worth noting that the US and UK are exceptions here. Our data also ends in 2019, before George Floyd in the US, the Casey Review in the UK, etc...

As you may have heard, our main finding is that trust in representative institutions has been declining globally (among democratic countries) but trust in implementing institutions has been stable or rising. But these broad strokes ofc. mask some interesting nuances.

Yeah when it's linear regression models; they're basically designed to provide predictions of y for any value of x. These are not just multiplications of the coefficients but also the differences between the predicted values in the first and last years derived from those models.