jueves, 10 de noviembre de 2016

Outline


For the purposes of the final paper, I will use the Ministry of Development and Social Inclusion as the organization that is trying to launch a reengineering process in which people could apply online for a recategorization of their poverty status in order to be eligible for social programs.
The structure of the paper will be the following:
·        The introduction:
Presentation on the organizations and the project’s objectives and goals and how they align with each other. This section will also cover how the current process is working now and how bringing it online could resolve it.
·        Defining the user and their needs:
The user are people living in poverty conditions. Although not all of them will have access to internet, it’s important to note that there are two stages in which the recategorization occurs. First, they apply in paper forms with their local office and then the office sends the information via email to the main office. So there are two stages at which the Ministry could intervene.

·        Articulating the value hypothesis: in order to test if the product or process is valuable to a consumer it will be important to analyze the benefits its creating to the target user in terms of time, money and other aspects of their life as well as the externalities both positive and negative.
·        Outlining an MVP
·        What hypothesis you have that need testing (A/B Testing)
·        The scale hypothesis
·        Culture/buy-in from within your organization
·        Opportunities and challenges around collecting and analyzing data

·        Final conclusion and recommendations

jueves, 27 de octubre de 2016

MAKE IT OR BREAK IT: The pains of accessing Peru's most respected Ministry's open data


I tried to download data from the most important Ministry in Peru: the Ministry of Economics and Finance. It is one of the few respectable places to work and it is known for staying at the frontier of transparency and technical correctness.
Entering the web portal to access public spending data is one of the most common reasons why people enter the Ministry's webpage. The Ministry itself has spent a lot of resources making it as user friendly as possible. The site's name is actually called "Friendly Consultation". Among some friends of mine, we used to call it "Zero Friendly Consultation". 

And here is why:

When you first access the portal you have to decide the module that you want to access Public Spending or Public Revenue, and if you want the data to be updated monthly or daily. It is still unclear what varies with the update, and they give no indication of the differences or trade-offs, so right off the bat, if you don't know the difference, it could alter whatever analysis you want to conduct.

Assuming you accurately decide the best module (I went for Public Spending with monthly update), you are then welcomed to the portal by one line of the total budget for the present year and different columns with different numbers. One point in favor of the site is that each line has a hover option where it tells you what each column mean, whether it's how much money was budgeted, modified budget, or actual spending, for example. Then you can access different levels of spending. 

On the surface it looks like it's going to be relatively easy. 

The next problem comes at this stage, you have to decide what level you want for a certain type of spending. For example, to obtain data from a social program you have to enter the Ministry of Development line of budget and then in spending subgroups and so on. The problem here is that people normally don't know how the public spending is structured and could be easily deceived or obtain the wrong information. 

Assuming you find exactly what you want, the most common tasks is wanting to compare the same type of expenditure for a given period. The problem here is that there is no way of selecting a range of years and downloading the data, you have to change each year every time and then download the spreadsheet. Moreover, and more scandalously, every time you change the year you have to start all over again, because it resets you search to the total budget! 

I believe this is because some budget lines switch positions over the years, still, I am sure there must be a more efficient way to show the changes without the user having to suffer so much. If the Ministry would dedicate time to actually making that consultation friendly, a lot of people would be more encouraged to analyze government public spending. The trade off would be finding weird stuff, but in reality people are already looking for that, so there is only much to gain from making it user friendly. With the new trend of pay for performance in budget allocation, the Ministry would benefit from making the user more engaged and interested in the data available

Again, another expectation v. reality moment. Here is a visualization on how the total budget has changed in the past 10 years. 


miércoles, 12 de octubre de 2016

Who are we protecting?


The Ministry of Development and Social Inclusion (MIDIS) is the newest Ministry in Peru, that has taken over almost all social programs since 2012. It is also in charge of the algorithm that decides who is eligible for social programs.
After my experience of downloading LastPass there are a few considerations that I would like MIDIS for take into account before mandating all of their employees to use it.

First, it is important to consider the positive externalities that it would generate to both the users and the Ministry itself. For the users/employees, it allows them to have their passwords stored in a "secure" way. There are countless times in which you just can't remember that one password you created with a certain pattern so its easy to remember but with a little twist so that it's different from the others. We all do it. (Did I just give up my password?). Anyway, it is nice to know that you have a safe place where all of that is stored and relieves the mind from that stressful moment and you are in a hurry. Second, I would say that then you can change your passwords and make them more intricate and complex therefore making them harder to guess/hack and by default more secure. Especially since you use a pattern, cracking one password will make it easier to hack the other ones. Third and last, you can use the google chrome plug and even the cellphone app in order to make it easier to access your accounts. The plug in even detects when you are accessing a site with credentials and asks if you would like to save it automatically, which makes the process much easier for the user.

For the Ministry, I think the main benefit is protecting their server from a possible hack. Imagine one of the employees had weak passwords or was very careless about how they manage it. It would expose the entire Ministry servers to stealing information. And take into account the fact that this Ministry is in charge of assessing who is eligible for the biggest social programs in the country. I think that alone is enough reason to encourage employees to add a layer of security to their access credentials.

On the other hand, let's consider the negative consequences that could happen from instating such policy. Both the users and the Ministry itself has to be aware of how LastPass actually works and what kind of security it has to protect the credentials of their users. If LastPass had a security breach it would mean exposing as much as information the employees has stored there, including banking access codes. Would that make the Ministry liable for a sue? Unclear at this point given that Peruvian Law is not as caught up with online services. However, it would generate social unrest and political liability to the incumbent party. The fact that LastPass is the platform that could unlock other platforms is extremely dangerous if it is hacked.

It is important to weigh in the trade offs between having your employees store their passwords in one place, especially if there is a security breach that could have big consequences . Even more so in Government. In conclusion, and just because I am risk averse, I would start by doubling down on securing the networks at the Ministry and engaging in better practices. Start small and then evaluate whether it is convenient to instate such a policy. It might not be the best option for all Ministries, or even for all employees in one Ministries. Try some adaptation if it is planned to roll-out and learn if it is really adding value and protecting the users and the Ministry from external threats.


miércoles, 21 de septiembre de 2016

Platforms in Government


Platforms in Government? Is that a thing?
I have to admit thinking about platforms is extremely hard when you don't have a "techy" background. It gets harder to do when you want to extrapolate the concept to government. And it gets way harder when you come from a developing country where a CTO position is reduced to making sure that the servers don't crash (still important but less impressive). 
I try to think of platforms as the necessary foundation for any given development. It's the concrete or abstract structure that supports any given activity. Trying to come up with an example, first thing that comes to mind when I first entered to work in the Ministry of Social Development and Inclusion was the process of applying. They post jobs online and their description, requirements, salary, division, etc. (I never understood why there were independent hiring offices across governments that did basically the same thing.) 
I encountered that in order to apply to a position I had to enter to the application page (a.k.a. platform), subscribe, obtain an user ID and password, and register. After registering, I had to input all of my education information and attach the corresponding documents, as well as my job experience with precise dates and, again, a letter from my former employer certifying the dates and position I had held. 
It was a very tedious process, especially asking that of my former employers. However, is a way to apply online to a job, reducing the cost of going to the Ministry during regular business days and waiting in line to drop an envelope with the physical legalized copies. From what I had heard it certainly was a way to improve the process of applying, which is very tedious in itself. You only had to go through that process if you had been selected for a position after several interviews and then you had to drop off the documents in order for them to verify it. If anything, it was certainly a way to waste less paper, and time. 
I had never thought of this online application as a platform. Especially in terms of human resources, I started wondering why was it that only this Ministry had it. And then I realized that probably the others had it too, but each division of the Government did. Then I thought about how each branch would've had to put out a procurement notice or try to develop it in house. I certainly don't understand why there isn't one single platform across government branches. I could easily come up with a lot of reasons why this save the Government enormous amounts of money. 
This set up definitely divided power between all hiring government branches. However, given that each branch handles their own evaluation of the proposals, I still find little use that the platform itself has to be done in house. Nonetheless, I do recognize the fact that security could be an issue, and if one of the centralized servers crashed it could be bad. There are thousands of hiring processes going on every day across Government agencies, and although delaying them one more day doesn't seem like the end of the world, it could affect greatly already slow-paced institutions. Another matter to take into account is security. Hacking all of the public workers information and also all of the people that have applied, could be done easily, especially taking into account the fact that the Peruvian Government is not very sophisticated with online security. 
Although there are pros and cons I still think even having just one available link that could direct you towards an application could make the process more streamlined. The risks of having a unified platform and preventable. Moreover, the cost of implementation would definitely save the Government a lot of in house operations. The economist in me can only see gains and efficiency. The public-policy-driven-HKS-student in me is very scared about implementing these types of things in developing countries: it sounds like a very typical case of premature load bearing for institutions. 
Expectation v. Reality, always.

miércoles, 14 de septiembre de 2016


Value Proposition Design: National Statistics Office in Peru


As an economist interested in social policy the Peruvian National Statistics Office webpage is a site I've used often since I was an undergrad back in Lima. It has provided resources of information used to calculate poverty rates, estimate demand in education, among many others tasks along my professional career. However, as a user that is constantly accessing the page I have noticed that there is steep learning curve into using this data. Only someone that has struggled several times to access it and knows which databases are available and differences between them can access it quickly. And so I chose this government service for my Value Proposition Design. 

Here is a snapshot of the introductory page to the data available. Once you navigate a little through the National Statistics Office website (INEI, in Spanish) it takes two clicks to arrive here. Notice, that there are two different tabs. One for documentation of the data and the other for the data itself. 



So here is how I thought the costumer profile would look like:

Customer Segment: Young economist working in development in Peru. 



And here is the Value Map:

As it can be expected: the needs of an Economist are very hard to please. 

sábado, 10 de septiembre de 2016

Right after finishing college I had an offer to work as a research assistant but heard of an offer to work as an intern for the very prestigious IPA in an education project. Two weeks in, I was offered a payed salary for working closely with Government for producing impact evaluation. The project was called “Comisión Quipu”. The goal of the CQ was to promote evidence-based policy through a series of meetings between highly respected local and international academics and government officials. The end goal project and these meetings was to execute the proposals to test a theory of change within the Social Programs.

It failed.

When the proposals were executed, I was to become the Project Associate for said evaluation.

It didn’t happen.


On paper the idea sounded great. Everyone was excited. IPA put a lot of resources both human and financial into making this project happen.

The first warning sign should’ve been how excited the Government was as well.

Our first obstacle when preparing for the meetings was when we found out that the Ministry had strong recommendations of who the academics should be. IPA had selected 6 highly respected PhDs but had to compromise 3 in order for there to be room for those recommended by the Ministry. The cost that it came with was having 3 people with long trajectory in working with the Government but had no idea what an impact evaluation was. This also implied that our team had to work harder than we thought because we had to start by giving a very brief course on impact evaluations and why they were important. Something that wouldn’t have been necessary in the first place.

Every week the academics had a deliverable, that started with pitching ideas to the government officials and getting feedback, and ended with the final report about the chosen evaluation. Given that some were not verse in impact evaluations we had to give a lot of feedback and comments to each document. More so than initially planned.

As the meetings started, it was increasingly clear that the Government was distancing itself from the CQ. They called it a good initiative or a starting point. They also said how the evaluations could be executed somewhere in the future. (Not the immediate future as we hoped.)

It is important to know that it was the year the Ministry had just been created by the new Government so their priorities were not focused on testing new ways of reducing poverty (although it should’ve been, in my opinion) but to roll out already thought out Program’s themselves and improve the process. They also were under the public eye because the Ministry had been one of the President’s promises during his campaign.

Now it seems pretty obvious that under those conditions the Government wouldn’t agree to start trying novel things right away, but this was a new type of project in IPA and in Peru so the steps that had to be taken were very unclear. It had never been done before and everyone got carried away with the novelty that no one stopped to try to assess how likely would it be for the Government to finance one of the evaluations.

The Government didn’t want to or couldn’t commit to funds to evaluate certain projects or to launch a pilot just because IPA went to all the trouble to bring international academics to sit with them and discuss ideas – including the founder of IPA. Especially because some of the things we wanted to test had never been implemented before, which made a new Government very uncomfortable. If they failed, they would’ve been greatly criticized. Maintaining the status quo of rolling out programs was much easier for them and less risky. At least that was my impression.

Looking back, it could’ve taken us probably two or three weeks to figure out that the CQ was not going to lead where we wanted it to. Given the political context, especially. But as I said, it had never been done before. IPA wanted to show that they were the leading organization in advocating for evidence-based policy with Governments. And the Government was trying to give an image of being a technically correct institution to its critics.

All that was left of CQ was the paper trail, and a valuable lesson for IPA and myself about how good intentions in Government are nothing more.  

You would think that after that, there would be no more CQ. You’d be wrong.
One year later, they did a second version of CQ. Why? I have no idea. What was the result? Same as before.

I have no (more) words.