Jonathan Haidt's "The Happiness Hypothesis" makes the point that very few things actually long-term moves the needle when it comes to happiness:
Meditation does. As does therapy. And drugs.
These can move our "normal range" of happiness. Most other things, Haidt argues, will temporarily bump our happiness up or down, but we quickly revert to our normal range.
Some things that generally don't affect our happiness in the long term according to Haidt: The amount of time we work or spend on leisure activities. Major health problems that are not degenerative (heart attack, losing your legs). Winning the lottery. Getting that big raise at work.
The reason, it seems, is that we adapt very quickly. Winning the lottery will make us very happy for a short while, but without a lot of discipline we've shortly adapted to our new-found wealth and the needle goes back down to our "normal range". Losing your legs will make you depressed for a while, but soon we adapt and the needle springs back up. Getting a raise is great for a month or three, but then we're used to it and looking forward to the next one.
As long as things are steadily getting better, we are mostly happy. If they are steadily getting worse, we are mostly unhappy. Even seemingly earth-shattering events that totally changes our lives only tends to affect most people for a while.
This does explain some of the obsession with GDP: It is a proxy for growth, and ongoing growth is associated with happiness. But as point out, optimizing for GDP can happen at the cost of negatively affecting other things that are also associated with happiness, and the net result is not necessarily all that great.
An interesting aspect of Haidt's claims is that we should not aim for a too rapid improvement if they make further improvements harder. That is, it's better to get a steady 3% increase in salary, than 50% now, and then nothing for many years. If you do get that 50% raise (or a huge exit), it is better not just for your finances but for your happiness to invest most of it, and slowly increase your spending over time.